Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
sequencelengths 1
91
| keywords
sequencelengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
Oncogenic signaling pathways regulate gene expression in part through epigenetic modification of chromatin including DNA methylation and histone modification . Trimethylation of histone H3 at lysine-27 ( H3K27 ) , which correlates with transcriptional repression , is regulated by an oncogenic form of the small GTPase Ras . Although accumulation of trimethylated H3K27 ( H3K27me3 ) has been implicated in transcriptional regulation , it remains unclear whether Ras-induced changes in H3K27me3 are a trigger for or a consequence of changes in transcriptional activity . We have now examined the relation between H3K27 trimethylation and transcriptional regulation by Ras . Genome-wide analysis of H3K27me3 distribution and transcription at various times after expression of oncogenic Ras in mouse NIH 3T3 cells identified 115 genes for which H3K27me3 level at the gene body and transcription were both regulated by Ras . Similarly , 196 genes showed Ras-induced changes in transcription and H3K27me3 level in the region around the transcription start site . The Ras-induced changes in transcription occurred before those in H3K27me3 at the genome-wide level , a finding that was validated by analysis of individual genes . Depletion of H3K27me3 either before or after activation of Ras signaling did not affect the transcriptional regulation of these genes . Furthermore , given that H3K27me3 enrichment was dependent on Ras signaling , neither it nor transcriptional repression was maintained after inactivation of such signaling . Unexpectedly , we detected unannotated transcripts derived from intergenic regions at which the H3K27me3 level is regulated by Ras , with the changes in transcript abundance again preceding those in H3K27me3 . Our results thus indicate that changes in H3K27me3 level in the gene body or in the region around the transcription start site are not a trigger for , but rather a consequence of , changes in transcriptional activity .
Epigenetic modification of chromatin is a key mechanism for regulation of gene expression [1] , [2] . Trimethylation of histone H3 at lysine-27 ( H3K27 ) is associated with transcriptional repression and is regulated by Polycomb repressive complex 2 ( PRC2 ) , a histone methyltransferase specific for H3K27 [3] . This modification of H3K27 ( H3K27me3 ) and Polycomb group proteins are thought to promote the formation of closed chromatin structures and thereby to repress transcription [4] , [5] . H3K27me3 controls Hox gene silencing and X chromosome inactivation , and it is therefore essential for normal development [6] , [7] . Dysregulation of H3K27me3 is also frequently observed in and is regarded as a hallmark of cancer , with global as well as site-specific increases or decreases in H3K27me3 levels having been detected in several tumor types [8]–[10] . Chromatin immunoprecipitation ( ChIP ) followed by deep sequencing ( ChIP-seq ) as well as chip-based ChIP have been applied to map precisely the distribution of H3K27me3 across the entire genome . These approaches have also been adopted to elucidate the relation between the distribution of H3K27me3 and transcriptional activity . Such studies have revealed at least two patterns of H3K27me3 enrichment associated with transcriptional repression: a focal enrichment around the transcription start site ( TSS ) and a broad enrichment encompassing the entire gene . H3K27me3 around the TSS frequently colocalizes with H3K4me3 and is associated with gene repression especially in undifferentiated cells [11] , [12] . A broad enrichment of H3K27me3 , also known as a blanket-type pattern or broad local enrichment ( BLOC ) , has been detected over larger genomic regions including the TSS [13]–[17] . This pattern of modification has been associated not only with individual repressed genes but also with repressed gene clusters , and it is frequently observed in differentiated cells . Furthermore , both of these enrichment patterns are highly variable among cell types [18] , [19] , indicating that the distribution of H3K27me3 is regulated in a manner dependent on the cellular and developmental context . The small GTPase Ras controls cell growth and survival in part through epigenetic modification including DNA methylation and histone modification . Ras regulates the activity of downstream signaling pathways including those mediated by mitogen-activated protein kinases ( MAPKs ) [20] , [21] . The activating G12V amino acid substitution is one of the most frequent Ras mutations found in human cancer . Ras up-regulates the expression of p16Ink4a , an inhibitor of cyclin-dependent kinases , and this effect is accompanied by a marked decrease in the amount of H3K27me3 at the Ink4a locus in mouse embryonic fibroblasts [22]–[25] . Moreover , Ras-induced oncogenic transformation of mouse NIH 3T3 cells is associated with the down-regulation of Fas , Reck , and Par4 transcription concomitant with an increase in DNA methylation [26]–[28] . Most of the reported associations between H3K27me3 status and transcription are based on correlation . It has thus remained to be determined definitively whether changes in H3K27me3 distribution are causal with regard to regulation of transcription . To elucidate the biological relevance of H3K27me3 , we have now investigated the time courses of Ras-induced changes in H3K27me3 level and in transcription at the genome-wide level in NIH 3T3 cells . Our results indicate that changes in H3K27me3 status follow , rather than precede , transcriptional changes induced by Ras signaling .
We established mouse NIH 3T3 cells that express a constitutively active mutant ( G12V ) of human H-Ras or that were infected with the corresponding empty retroviral vector ( referred to hereafter as Ras cells and Vec cells , respectively ) . Expression of the Ras transgene resulted in increased phosphorylation of the MAPK isoforms Erk1 and Erk2 ( Figure 1A ) as well as in morphological transformation of the cells ( Figure 1B ) . Moreover , reverse transcription ( RT ) and quantitative polymerase chain reaction ( qPCR ) analysis revealed that the Ras cells exhibited transcriptional repression of Fas locus genes including Fas , Acta2 , and Stambpl1 ( Figure 1C ) , consistent with previous observations [26] , [28] . Fas is a cell surface receptor that mediates the induction of apoptosis by Fas ligand [29] . Although Ras signaling has been reported to increase the level of DNA methylation around the Fas locus [26] , [28] , we did not detect such an obvious effect ( data not shown ) . To identify histone modifications that might contribute to silencing of the Fas locus , we performed ChIP-qPCR analysis with antibodies to transcriptionally repressive histone marks including H3K9me2 , H3K9me3 , and H3K27me3 ( Figure 1D ) . Among these marks , only the amount of H3K27me3 was increased at the Fas locus of Ras cells . The H3K27me3-enriched region contained the entire Fas gene as well as the promoter of Acta2 . These results thus showed that Ras signaling induces trimethylation of H3K27 as an epigenetic modification . To determine whether Ras-induced changes in H3K27me3 abundance are a trigger for or a consequence of changes in transcription , we set out to analyze the time courses of these events at the genome-wide level in NIH 3T3 cells infected with the retrovirus for H-Ras ( G12V ) at time 0 . Transcript and H3K27me3 levels were measured by RNA-seq and ChIP-seq , respectively ( detailed sequencing information is provided in Table S1 ) . First , we identified regions of H3K27me3 enrichment associated with silent genes in cells before introduction of H-Ras ( G12V ) ( Ras0 cells ) ( Figure 2A ) . H3K27me3 showed broad enrichment domains encompassing several hundred kilobases , consistent with previous observations [13] , [15] . To characterize the pattern of H3K27me3 within genes , we divided each gene into the gene body , upstream region , and downstream region , with gene body being defined as the genomic region from the TSS to the transcription termination site ( TTS ) . From a total of 23 , 232 RefSeq genes , we randomly selected 2000 genes and ordered them according to similarity in the pattern of H3K27me3 enrichment ( Figure 2B ) . This analysis revealed that the pattern of H3K27me3 enrichment fell into three distinct clusters ( designated brown , gray , and purple clusters ) . In the gray cluster , H3K27me3 covered the gene body as well as the region around the TSS . This cluster contained a high proportion of transcriptionally repressed genes , as represented by the bluish color in the FPKM ( fragments per kilobase of exon model per million mapped fragments ) column . This finding was confirmed by a different method examining all RefSeq genes , as detailed below . We next focused on the H3K27me3 signal in the gene body or in the region around the TSS of each gene ( Figure 2C ) . RefSeq genes were classified into five groups according to their expression level . In the groups containing repressed genes ( FPKM of 0 or 0–1 ) , H3K27me3 was localized to the gene body as well as to the region around the TSS ( Figure 2D ) . In contrast , in the groups containing expressed genes ( FPKM of 1–10 , 10–100 , or >100 ) , H3K27me3 was present at a low level in the gene body and in the nucleosome-free region around the TSS . The mean H3K27me3 signals in the gene body and in the region around the TSS of each gene also reflected the transcriptional status of the corresponding genes ( Figure 2E ) . These data indicated that enrichment of H3K27me3 in the gene body as well as in the region around the TSS reflects silenced transcription . We next identified genes whose transcription and H3K27me3 level are both regulated by Ras . We calculated the fold change in mean H3K27me3 level over the gene body for individual genes in cells infected with the Ras retroviral vector for 2 , 4 , 7 , or 12 days relative to that in Ras0 cells . Among a total of 23 , 232 RefSeq genes , 1027 genes showed at least a twofold change in H3K27me3 level at least one time point ( Figure 3A ) . A total of 933 genes showed a significant change in expression level at at least one time point after Ras introduction ( see Materials and Methods ) . We then subjected the 115 genes whose H3K27me3 level and expression were both regulated by Ras to hierarchical clustering based on the time course of the change in H3K27me3 abundance ( Figure 3B ) . This analysis revealed three distinct patterns of H3K27me3 dynamics induced by Ras: A purple cluster of genes in which the H3K27me3 level increased after Ras activation , and gray and brown clusters in which the H3K27me3 level decreased . Whereas changes in H3K27me3 abundance in the brown cluster were not associated with a characteristic transcriptional trend , those in the purple and gray clusters were inversely correlated with changes in transcription ( Figure 3B , Figure S1A and S1B ) . Moreover , in these transcription-correlated clusters , changes in transcription were apparent within 2 days after Ras activation , whereas the mean H3K27me3 level remained essentially unchanged at this time point ( Figure 3C ) . We calculated “t-half” to evaluate the timing of these two events ( Figure S1C ) . In the purple cluster , the median t-half for mRNA abundance occurred at 1 . 1 days and that for H3K27me3 level occurred at 6 . 9 days ( Figure 3D ) . In the gray cluster , the median t-half for mRNA abundance occurred at 3 . 3 days and that for H3K27me3 level occurred at 4 . 8 days . These results thus indicated that changes in transcription precede those in H3K27me3 level in the gene body . We performed a similar analysis for the 196 genes whose H3K27me3 level in the region around the TSS and expression were both regulated by Ras ( Figure S2 ) . Similar to the case for H3K27me3 in the gene body , increases in H3K27me3 level in the region around the TSS occurred after decreases in transcription . Together , our genome-wide comprehensive analyses thus revealed that Ras signaling affects transcription before it affects mean H3K27me3 level both in the gene body and in the region around the TSS . We selected three gene loci—Itgb5 , Adcy7 , and Smad6—for further study to confirm the results of our genome-wide RNA-seq and ChIP-seq analyses . Itgb5 and Adcy7 manifested Ras-induced changes in H3K27me3 level in the gene body ( Figure 4A ) . The time courses of the ChIP-seq and RNA-seq data showed that Ras signaling initially affected transcription and then gradually changed the H3K27me3 content of the gene body for Itgb5 and Adcy7 ( Figure 4B ) as well as for four additional genes , Plekha4 , Ephx1 , Bpifc , and Sorcs2 ( Figure S3 ) . In the case of Smad6 , the H3K27me3 level increased prominently in the region around the TSS but only slightly in the gene body as previously reported ( Figure 4A ) [30] . Ras signaling again affected transcription first and then gradually changing H3K27me3 content ( Figure 4B ) . In addition to Smad6 , we found other genes that showed a prominent increase in H3K27me3 level in the region around the TSS by visual inspection of the genome browser ( data not shown ) . These results for Itgb5 , Adcy7 , and Smad6 were confirmed by RT-qPCR and ChIP-qPCR analyses ( Figure 4C ) . We also confirmed that changes in gene expression precede those in H3K27me3 level with the use of NIH 3T3 cells that stably express Raf-ER , a fusion protein composed of the catalytic domain of Raf-1 and the ligand binding domain of the estrogen receptor . Treatment of these cells with 4-hydroxytamoxifen ( 4HT ) activates Raf-ER and downstream MAPK pathways [31] . Activation of Raf-ER thus also affected mRNA abundance before H3K27me3 level for Itgb5 , Adcy7 , and Smad6 ( Figure 4D ) as well as for four additional genes , Plekha4 , Ephx1 , Bpifc , and Sorcs2 ( Figure S7A ) . We also evaluated the Ras-induced changes in transcription and H3K27me3 level in the gene body for Itgb5 and Adcy7 by independent deep sequencing and qPCR with several primer sets ( Figure S4 ) , again confirming our results . Total histone H3 level in the gene body of Itgb5 or Adcy7 was affected only slightly by Ras signaling ( Figure S4D and S4H ) . The altered H3K27me3 content of the gene body was thus likely due to a change in H3K27 trimethylation , not to a change in nucleosome density . In addition to H3K27me3 , we also examined H3K9me2 and H3K9me3 levels ( Figure S5 ) . Among these repressive histone marks , only H3K27me3 was markedly altered by Ras signaling . Together , these data suggested that our genome-wide analyses correctly identified genes that undergo changes in transcription and H3K27me3 level in response to Ras signaling , and they confirmed that the changes in transcription precede those in H3K27me3 level . Our results suggested that a change in the amount of H3K27me3 is not required for Ras-induced regulation of gene transcription . To verify this hypothesis , we prepared NIH 3T3–Raf-ER cells depleted of H3K27me3 by transfection with small interfering RNAs ( siRNAs ) for Suz12 , a subunit of PRC2 that is indispensable for methyltransferase activity at H3K27 [32] . The cells were transfected with Suz12 siRNA for 48 h before exposure to 4HT for 24 h ( Figure 5A ) , and they were then analyzed for effects on H3K27me3 and transcription . Immunoblot and ChIP-qPCR analyses revealed that knockdown of Suz12 resulted in depletion of H3K27me3 in the total chromatin fraction ( Figure 5B ) as well as at specific regions such as Itgb5 , Adcy7 , and Smad6 loci ( Figure 5C , Figure S6A–S6C ) . Depletion of H3K27me3 did not affect the 4HT-induced repression of Itgb5 and Smad6 expression ( Figure 5D ) , indicating that an increase in the level of H3K27me3 is not required for Ras-induced transcriptional silencing of these genes . Furthermore , H3K27me3 depletion did not induce expression of Adcy7 in the absence of 4HT ( Figure 5D ) , indicating that depletion of H3K27me3 is not sufficient to induce transcriptional activation . We obtained similar results with two additional Suz12 siRNAs ( Figure S6D–S6F ) and four additional genes , Plekha4 , Ephx1 , Bpifc , and Sorcs2 ( Figure S7 ) . We also examined the effect of H3K27me3 depletion after the activation of Raf signaling by transfecting NIH 3T3–Raf-ER cells with Suz12 siRNA 3 days after exposure to 4HT ( Figure 5E and 5G ) . Analysis of the cells at 7 days after the onset of Raf activation revealed that Suz12 siRNA efficiently suppressed the increase in H3K27me3 level at Itgb5 and Smad6 ( Figure 5F ) . Nevertheless , this effect did not induce expression of Itgb5 and Smad6 ( Figure 5G ) , indicating that depletion of H3K27me3 does not affect transcriptional suppression of Itgb5 and Smad6 by Ras signaling . These data suggested that changes in H3K27me3 abundance do not play a critical role in the induction of gene silencing at later stages of Ras activation . Together , our observations indicated that a change in the level of H3K27me3 induced by Ras is not a trigger for , but rather a consequence of , a change in transcription . The presence of H3K27me3 at an exogenous transgene was previously shown to maintain the repressed state [33] , suggesting the possibility that an increase in H3K27me3 level induced by Ras signaling might be able to maintain repression of gene expression after signaling is inactivated . To test this possibility , we introduced ER-Ras [a fusion protein of human H-Ras ( G12V ) and the estrogen receptor] into NIH 3T3 cells , exposed the cells to 4HT for 9 days in order to induce changes in both H3K27me3 level and transcription , and then removed 4HT to inactivate Ras signaling ( Figure 6A ) . Immunoblot analysis revealed that ER-Ras was induced by 4HT and that its abundance decreased rapidly after removal of 4HT ( Figure 6B ) , the latter indicative of inactivation of the Ras signal . Changes in the transcription of Itgb5 , Adcy7 , and Smad6 were also apparent after exposure of the cells to 4HT for 9 days , whereas these changes were completely reversed after 4HT removal ( Figure 6C ) . Moreover , an increase in H3K27me3 content at Itgb5 and Smad6 was observed in the presence of 4HT , whereas H3K27me3 abundance at these genes returned essentially to basal levels after signal inactivation ( Figure 6D ) . The H3K27me3 level at Adcy7 was reduced by exposure of the cells to 4HT and remained low after 4HT removal , suggesting that the dynamics of H3K27 methylation and demethylation might differ . We obtained similar results with four additional genes—Plekha4 , Ephx1 , Bpifc , and Sorcs2 ( Figure S8 ) —as well as with cells expressing Raf-ER ( Figure 6E–6G ) . From these data , we concluded that changes in H3K27me3 level are dependent on Ras signaling , and that H3K27me3 enrichment is not maintained after inactivation of such signaling , resulting in reactivation of transcription . Visual inspection of H3K27me3 distribution revealed that Ras signaling alters H3K27me3 levels in intergenic regions located several kilobases distant from known gene bodies . Two representative loci , Col1a1 and Mink1 , are shown in Figure 7A . H3K27me3 was enriched in the region upstream of Col1a1 but was depleted in the region upstream of Mink1 in Ras cells . Given that changes in H3K27me3 level were frequently observed in the transcribed region of genes such as Itgb5 and Adcy7 , we examined whether unannotated transcripts might be produced from the regions upstream of Col1a1 and Mink1 . We reanalyzed RNA-seq data obtained by SOLiD sequencing , which contain strand information ( see Materials and Methods ) , and we indeed detected sequence reads for these regions , suggesting the existence of corresponding transcripts ( Figure 7A ) . RT-qPCR analysis confirmed the presence of transcripts derived from the regions upstream of Col1a1 and Mink1 ( hereafter referred to as uCol1a1 and uMink1 , respectively ) ( Figure 7B ) . Ras signaling repressed uCol1a1 expression and activated uMink1 expression , similar to its effects on Col1a1 and Mink1 mRNA levels ( Figure 7A ) . These results thus revealed that changes in the H3K27me3 content of intergenic regions can predict the presence of unannotated transcripts . To determine whether the changes in uCol1a1 and uMink1 transcription also precede those in H3K27me3 level , we examined the respective time courses with cells expressing Raf-ER ( Figure 7C ) . The expression of uCol1a1 and uMink1 was altered already at 1 day after exposure of the cells to 4HT , whereas H3K27me3 level remained essentially unaffected at this time . The level of H3K27me3 changed at 5 days ( uCol1a1 ) or 3 days ( uMink1 ) after 4HT exposure . We obtained similar results for transcripts derived from another intergenic region , uIl33 ( Figure S9 ) . Ras-induced transcription from intergenic regions thus also occurs prior to changes in H3K27me3 level . To examine whether H3K27me3 is required for regulation of uCol1a1 and uMink1 transcription , we determined the effect of Suz12 knockdown with siRNAs . Depletion of H3K27me3 did not affect the 4HT-induced silencing of uCol1a1 , nor did it induce uMink1 expression in the absence of 4HT ( Figure 7D ) . Together , these results suggested that the observed changes in H3K27me3 level in transcribed regions result from changes in transcription .
H3K27me3 has been found to manifest at least two distinct enrichment patterns—being abundant in narrow regions around the TSS and in broad domains that include entire genes—and the appearance rate of these patterns differs among cell types [18] . We have now analyzed these patterns in control NIH 3T3 ( Ras0 ) cells . For a simple comparison of H3K27me3 level with transcription in cells at various times after the onset of Ras expression , we used the mean value of H3K27me3 level in a defined region such as the gene body or the region around ( ±2 . 5 kb ) the TSS to represent the H3K27me3 status of each gene ( Figure 2C ) . We found that enrichment of H3K27me3 not only in the region around the TSS but also in the gene body correlated inversely with transcriptional level in these cells , consistent with previous observations [18] . Comparison of the time courses of mean H3K27me3 level and transcription allowed us to identify genes for which H3K27me3 content changes together with transcriptional activity in response to Ras signaling , suggesting that the mean value of H3K27me3 level in the defined regions provides an indication of H3K27me3 status of individual genes under different cellular conditions . We noticed by visual inspection the existence of several patterns of H3K27me3 modification within the defined regions . Although the use of mean values of H3K27me3 level disregarded these patterns , we conclude that such mean values provide a relatively simple measure for comparison of H3K27me3 status with transcriptional activity . For example , Itgb5 manifested a typical broad increase in H3K27me3 level , whereas Adcy7 showed two discontinuous regions of H3K27me3 enrichment in the gene body that were depleted in parallel in response to Ras signaling , and Smad6 exhibited a prominent increase in H3K27me3 around the TSS ( Figure 4A ) . Although various internal patterns of H3K27me3 were observed , however , visual inspection revealed that the time courses of H3K27me3 level at each position in the defined regions were similar to those for the mean value ( Figure 4A and 4B ) , indicating that changes in mean H3K27me3 level in the defined regions also represent changes in H3K27me3 status of genes despite differences in the internal patterns within the defined regions . Our approach based on mean H3K27me3 level in defined regions thus allows evaluation of the timing of changes in H3K27me3 abundance relative to those in transcription , and it leads us to the conclusion that Ras-induced changes in H3K27me3 level occur after those in transcription . Our H3K27me3 ChIP-seq data contain time course information as well as higher positional resolution compared with previously published H3K27me3 ChIP-seq results [11] , [15] , [30] . Our data are thus amenable to analysis of other aspects of H3K27 trimethylation . For example , temporal analysis of H3K27me3 distribution at base-pair resolution might allow the unveiling of Polycomb response elements ( PREs ) , for which little information is currently available in mammal [36]–[38] . Our results show that the Ras signaling–induced changes in transcription precede those in H3K27me3 level . Previous studies have also shown that transcriptional regulation is initiated before changes in H3K27me3 content [39]–[41] . We further revealed that an increase in H3K27me3 level induced by Ras is insufficient for maintenance of transcriptional repression after inactivation of Ras signaling . Such increases in H3K27me3 level induced by Ras signaling were thus found to be completely reversed after signal inactivation . Similar reversibility of changes in H3K27me3 level has been described for the FLC gene in Arabidopsis [41] , for which transcription and H3K27me3 content are regulated by signaling that is responsive to changes in temperature . It is thus possible that a signal-induced increase in H3K27me3 abundance is dispensable for both initiation and maintenance of transcriptional repression in various cell types and different species . On the other hand , H3K27me3 has been reported to contribute to maintenance of transcriptional suppression in other experimental systems [33] , [42] . The combination of H3K27me3 with other epigenetic marks has also been found to be related to transcriptional repression [6] , suggesting the possibility that Ras might regulate only H3K27me3 , and not other histone marks required for maintenance of gene silencing . One such possible histone modification is ubiquitylation of histone H2A at lysine-119 [43] . H3K27me3 recruits PRC1 , which functions as a ubiquitin ligase for this residue of H2A . Ubiquitylation of H2A by PRC1 results in repression of transcription by blocking the release of RNA polymerase II from promoters [44] . Not all genomic regions that show H3K27me3 enrichment colocalize with ubiquitylated H2A ( H2Aub ) or PRC1 [45] , [46] , however , suggesting that both H3K27me3 and H2Aub may be required for maintenance of gene silencing . Given that Ras-induced changes in H3K27me3 level are a consequence of those in transcription , Ras might influence H3K27me3 content without affecting H2Aub level . Further analysis of H2Aub level during Ras activation may provide insight into the function of H3K27me3 . The mechanism by which Ras signaling regulates H3K27me3 level in NIH 3T3 cells remains unclear . Changes in histone modification are mediated by changes in the expression or localization of the corresponding enzymes [47] . Changes in the expression level of enzymes have thus been found to be responsible for changes in H3K27me3 level in response to Ras signaling [24] , [25] . We found that the expression of genes encoding subunits of PRC2 or PRC1 was not altered by Ras activation in NIH 3T3 cells , however ( Figure S10A and S10B ) . Of genes for two known demethylases , the expression of only Jmjd3 was found to be up-regulated by Ras signaling , consistent with previous observations [24] , [25] . However , knockdown of Jmjd3 expression did not affect the expression of Adcy7 in the absence or presence of Ras signaling ( Figure S10C ) , suggesting that the change in Jmjd3 expression level is not required for Ras-induced changes in transcription . Phosphorylation of several sites of Ezh2 by various kinases has been shown to alter the localization of PRC2 [48] , [49] . Moreover , Msk1 and Msk2 , which are downstream kinases of Ras phosphorylate serine-28 of histone H3 ( a residue adjacent to K27 ) and this phosphorylation prevents PRC2 from recognizing H3K27 and results in passive H3K27me3 demethylation during subsequent progression of the cell cycle [50] , [51] . Such phosphorylation might contribute to the regulation of H3K27me3 level by Ras in our system . It is also possible that RNA polymerase II actively erases H3K27me3 by recruiting an H3K27me3 demethylase to the transcribed region , as previously proposed [39] . In support of this idea , we found that the demethylated regions partially coincide with the gene body , along which RNA polymerase II moves . Moreover , we detected unannotated transcripts derived from intergenic regions whose H3K27me3 level is regulated by Ras . These findings indicate that transcription might trigger H3K27me3 regulation and determine the localization of H3K27me3 demethylases . Detailed analysis of H3 modification and the localization of these enzymes may provide insight into the mechanisms determining the specificity of genomic regions subject to changes in H3K27me3 level . We have found that Ras-induced changes in transcription precede those in H3K27me3 level , suggesting that transcriptional regulation by Ras is initiated by a mechanism independent of H3K27me3 . We also performed ChIP-qPCR analysis of active histone modifications and observed changes in acetylation of H3 that were coincident with initiation of transcriptional changes at 2 days after Ras induction ( data not shown ) . Removal of active histone marks by Ras is thus a possible mechanism for Ras-mediated gene silencing . In this case , the repressive H3K27me3 mark might be deposited passively on repressed genes as a result of the loss of acetylation . Further genome-wide and time course analyses of histone acetylation are required to examine this possibility .
NIH 3T3 cells were obtained from American Type Culture Collection ( CRL-1658 ) and were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum , 1% penicillin-streptomycin , 2 mM l-glutamine , 1% MEM–non essential amino acids , and 1% sodium pyruvate ( all from Life Technologies , Foster City , CA ) . Complementary DNAs encoding human H-Ras ( G12V ) or a fusion protein of human Raf-1 and the estrogen receptor ( Raf-ER ) were subcloned into the retroviral vector pMX-puro [52] . A pLNCX2 vector encoding a fusion protein of the estrogen receptor and human H-Ras ( G12V ) ( ER-Ras ) was kindly provided by M . Narita [53] . These vectors were introduced into Plat-E packaging cells by transfection with use of the FuGENE6 reagent ( Promega , Madison , WI ) . Culture supernatants containing recombinant ecotropic retroviruses were harvested for infection of proliferating NIH 3T3 cells in the presence of polybrene . The infected cells were then subjected to selection in medium containing puromycin for pMX-puro or G418 for pLNCX2 . Activation of Raf-ER or ER-Ras was induced by exposure of cells to 10 or 100 nM 4HT ( Sigma , St . Louis , MO ) , respectively , that had been dissolved in ethanol; the medium supplemented with 4HT was refreshed every day . Cells were transfected with Suz12 , Bmi1 , Jmjd3 , or control Stealth RNAi duplexes ( Life Technologies ) with the use of a Neon Transfection System ( Life Technologies ) . The Suz12 siRNA sequences are 5′-UAAAUUCUCUUCUUCCUGGACGAGU-3′ , 5′-UUUGAUUGAGGUCAGGAUUCAAAGG-3′ , and 5′-UAUCGUUGGUUUCUCCUGUCCAUCG-3′ for #1 , #2 , and #3 , respectively . The Bmi1 siRNA sequence is 5′-CGUCAUGUAUGAAGAGGAACCUUUA-3′ , and the Jmjd3 siRNA sequence is 5′-GGAUGACCUCUAUGCGUCCAAUAUU-3′ . Cells were lysed in a solution containing 50 mM Tris-HCl ( pH 7 . 6 ) , 300 mM NaCl , 0 . 5% Triton X-100 , aprotinin ( 10 µg/ml ) , leupeptin ( 10 µg/ml ) , 1 mM phenylmethylsulfonyl fluoride , 400 µM Na3VO4 , 400 µM EDTA , 10 mM NaF , and 10 mM sodium pyrophosphate . The lysate was centrifuged at 20 , 000× g for 10 min at 4°C , and the resulting supernatant was isolated as a cytosolic fraction . The pellet was resuspended in lysis solution for use as a chromatin fraction . Proteins in each fraction were resolved by SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane ( Millipore , Billerica , MA ) . Immunoblot analysis was performed with antibodies to H-Ras ( sc-520; Santa Cruz Biotechnology , Santa Cruz , CA ) , to Erk1/2 ( 9102; Cell Signaling Technology , Beverly , MA ) , to phosphorylated Erk1/2 ( 9101; Cell Signaling Technology ) , to Suz12 ( ab12073; Abcam , Cambridge , MA ) , to α-tubulin ( T5168; Sigma ) , to H3K27me3 ( 07-449; Millipore ) , and to histone H3 ( ab1791; Abcam ) . Immune complexes were detected with horseradish peroxidase–conjugated secondary antibodies and Super Signal West Dura Luminol/Enhancer Solution ( Thermo Scientific , Rockford , IL ) . The chemiluminescence signals were quantitated with a digital imaging system ( VersaDoc; Bio-Rad , Hercules , CA ) . Total RNA was isolated from cells and purified with the use of an SV Total RNA Isolation System ( Promega ) . It was then subjected to RT with the use of a PrimeScript RT reagent kit ( Takara Bio , Shiga , Japan ) followed by real-time PCR analysis with a StepOnePlus Real Time PCR System ( Life Technologies ) and Fast SYBR Green Master Mix ( Life Technologies ) . Data were analyzed according to the 2−ΔΔCt method and were normalized by the amount of acidic ribosomal phosphoprotein P0 ( Arbp ) mRNA . The sequences and gene information of PCR primers are listed in Table S2 . Cells were fixed with 0 . 6 or 1 . 0% formaldehyde for 10 or 5 min , respectively , at room temperature , after which glycine was added to the medium . The cells were then lysed and stored at −80°C until analysis . The lysates were thawed and subjected to ultrasonic treatment with the use of a Bioruptor ( Diagenode , Denville , NJ ) or Covaris S2 ( Covaris , Woburn , MA ) instrument in order to obtain chromatin fragments of 200 to 700 bp . Antibodies to H3K27me3 ( 07-449; Millipore ) , to H3K9me2 ( ab1220; Abcam ) , to H3K9me3 ( ab8898; Abcam ) , or to H3 ( ab1791; Abcam ) , or normal mouse ( sc2025; Santa Cruz Biotechnology ) or rabbit ( sc2027; Santa Cruz Biotechnology ) immunoglobulin G , were incubated with Protein A Dynabeads or Protein G Dynabeads ( Life Technologies ) to allow formation of bead-antibody complexes . Chromatin fragments were then subjected to immunoprecipitation with the bead-antibody complexes , after which the beads were washed and immunoprecipitated chromatin fragments were eluted and treated with RNase A and proteinase K . DNA was extracted from the samples with phenol-chloroform and was then precipitated with ethanol and dissolved in TE buffer . Quantitative PCR analysis of ChIP DNA was performed as described above . Primer sequences and positions are listed in Table S3 . Data were analyzed according to the 2− ( Ct of IP sample – Ct of input sample ) method and are presented as a percentage of input . For comprehensive analysis of Ras-dependent changes in gene expression and H3K27me content , we performed RNA-seq and ChIP-seq analyses at various times after Ras induction . We sampled cells at 0 , 2 , 4 , 7 , and 12 days after infection with the H-Ras ( G12V ) retroviral vector . We sequenced five and six samples for analysis of gene expression and H3K27me3 , respectively . ChIP-seq libraries were prepared from ∼40 ng each of ChIP and input DNA with the use of a TruSeq DNA LT Sample Prep Kit ( Illumina , San Diego , CA ) . RNA-seq libraries were prepared from 2 µg of total RNA with the use of a TruSeq RNA Sample Prep Kit v2 ( Illumina ) . Two flow cells ( 16 lanes ) of an Illumina HiSeq 2000 instrument were used . Libraries were clonally amplified in a flow cell and sequenced with the use of HiSeq Control Software 1 . 5 ( Illumina ) and a 48-nucleotide paired-end sequence . Image analysis and base calling were performed with the use of Real Time Analysis ( RTA ) 1 . 13 software . A total of 81 , 877 , 304 ( RNA-seq ) or 1 , 068 , 022 , 370 ( ChIP-seq ) reads was obtained per sample . For SOLiD sequencing , ChIP-seq libraries were prepared from 20 ng each of ChIP and input samples with the use of a SOLiD Fragment Library Construction Kit with SizeSelect Gels ( Life Technologies ) . For RNA-seq , total RNA ( 10 µg ) was subjected to rRNA depletion ( RiboMinus Eukaryote Kit for RNA-seq , Life Technologies ) and RNA-seq library construction ( SOLiD Whole Transcriptome Analysis Kit , Life Technologies ) . The libraries were clonally amplified on SOLiD P1 DNA Beads by emulsion PCR and sequenced with the SOLiD3Plus System ( Life Technologies ) to generate 50-base single-end reads . Sequencing data of ChIP-seq and RNA-seq are available under the accession number of DRA001075 from DNA Data Bank of Japan Sequence Read Archive ( DRA ) . FastQC ( http://www . bioinformatics . babraham . ac . uk/projects/fastqc ) analysis revealed low sequence quality for the last 4 bases of the second read of paired-end reads , and so these bases were trimmed before data analysis . For the sequence data analysis , UCSC mm9 and RefSeq were used as the reference mouse genome and gene model , respectively . For gene expression analysis , paired-end reads were mapped to the mouse genome with the use of TopHat ( ver . 2 . 0 . 8 ) [54] . Cufflinks ( ver . 2 . 0 . 10 ) [55] was used to estimate gene expression level on the basis of fragments per kilobase of exon model per million mapped fragments ( FPKM ) . Gene expression level was compared between control ( Ras0 ) cells and Ras cells at 2 , 4 , 7 , or 12 days after activation of Ras signaling with the use of Cuffdiff ( ver . 2 . 0 . 10 ) . A Q-value of <0 . 05 was set as a threshold for differential expression , resulting in the extraction of 933 genes as differentially expressed genes . For H3K27me3 analysis , sequenced reads were mapped to the mouse genome with the use of bwa ( ver . 0 . 5 . 9 ) [56] . Paired reads that were uniquely mapped to the genome were extracted . This filtering process discarded 174 , 644 , 447 ( 16 . 72% ) reads per sample , and the remaining 870 , 116 , 495 ( 83 . 28% ) reads were used for subsequent analyses . The sequence depth of our ChIP-seq data set was 40 ( including insert ) , with 90% of bases in the genome being covered by at least one read . As far as we are aware , this is one of the most deeply sequenced histone modification marks to date . For the purposes of our study , we defined the gene body as the genomic region from the TSS to the TTS . Introns are thus included in the gene body as well as exons . To compare methylation signals associated with genes of different sizes , we calculated the relative position of bins in the gene body as follows: ( 1 ) where gs and ge represent the TSS and TTS of gene g , bis and bie are the start and end positions of the ith bin in gene g , and ris and rie are the normalized ith bin positions , respectively . The same normalization procedure was applied to the upstream and downstream regions of each gene , resulting in 600 data points per gene . The methylation signal associated with each bin was then assigned to the normalized positions . The mean and standard deviation of the methylation signal for the 600 data points obtained for each gene were calculated . The methylation signal around the TSS ( TSS ±2 . 5 kb ) was analyzed in the same manner . To examine the methylation pattern across the gene body and adjacent regions , we randomly selected 2000 out of 23 , 232 RefSeq genes . The methylation signal of each selected gene was normalized so that the mean and standard deviation of the signal were equal to 0 and 1 , respectively . A hierarchical clustering was performed according to the normalized methylation signal , and the results were visualized as a heat map in which the intensity of the methylation signal was color coded . Gene expression level was also displayed in the heat map ( Figure 2B ) . The relation between methylation and gene expression was investigated by comparison of the average methylation signal and FPKM . For this purpose , the average methylation signal in the gene body and in the region around the TSS ( defined here as the region from −2 . 5 to +2 . 5 kb relative to the TSS ) was calculated for each gene at all time points . Gene expression level was categorized into five classes including unexpressed genes ( FPKM = 0 ) , and the distribution of the methylation signal in each class is presented as a box-and-whisker plot ( Figure 2E ) . The average methylation signal in control ( Ras0 ) cells compared with cells at various times after infection with the H-Ras ( G12V ) retroviral vector as well as the fold change in the methylation signal were calculated . If the latter fold change was ≥2 , then the gene was extracted as a differentially methylated gene . A total of 1027 genes ( gene body ) or 1230 genes ( region around the TSS ) fulfilled this criterion , and these genes were further examined by comparison with differentially expressed genes . To examine changes in methylation over time , we performed a hierarchical clustering according to average methylation signal as described above ( Figure 3B , Figure S2B ) . To examine whether gene expression might be causally related to a change in H3K27me3 level , we defined and calculated “t-half” as shown in Figure S1C . Two such values were calculated for each gene , one for gene expression and the other for H3K27me3 level . We obtained 130 , 133 , 653 , 150 , 914 , 422 , 226 , 490 , 377 , and 250 , 941 , 002 reads from control ( Vec ) and Ras cells for RNA-seq and from Vec and Ras cells for ChIP-seq , respectively . Sequenced reads were mapped to the mouse genome with the use of the BioScope Map Data program ( ver . 1 . 2 ) . The following analyses were based on the mapping results . Gene expression level was estimated by calculating reads per kilobase of exon model per million mapped reads ( RPKM ) [57] . Overall gene expression level in Vec and Ras cells was normalized by the expression level of Arbp . To calculate the amount of H3K27me3 , we split the mouse genome into 1-kb bins and used the number of reads in each bin as the raw methylation signal . This raw signal was normalized as Vec and Ras cells have the same number of reads and was then smoothed with the lowess function . Obtained signals were converted into wiggle ( WIG ) format and uploaded to the UCSC Genome Browser for visualization . To identify putative novel transcripts , we mapped sequenced reads to the mouse genome with the use of TopHat ( ver . 1 . 3 . 3 ) . Given that our SOLiD sequence data included strand information , mapped reads on the Watson and Crick strand were analyzed separately . The number of mapped reads at each genome coordinate was converted to bigWig format and uploaded to the UCSC Genome Browser . Those regions that did not overlap with a RefSeq gene and showed a difference in expression level between Vec and Ras cells were manually inspected . | Trimethylation of histone H3 at lysine-27 ( H3K27 ) has been associated with silencing of gene expression . Abnormalities of this modification are thought to contribute to the epigenetic silencing of tumor suppressor genes and are regarded as a hallmark of cancer . It has remained unclear , however , whether the production of trimethylated H3K27 ( H3K27me3 ) is the cause or the consequence of gene silencing . To address this issue , we examined the time courses of changes in H3K27me3 level and those in gene transcription induced by an oncogenic form of the Ras protein , the gene for which is one of the most frequently mutated in human cancer . We found that the amount of H3K27me3 was inversely related to transcriptional activity both at the genome-wide level and at the level of individual genes . However , we also found that the Ras-induced changes in H3K27me3 level occurred after those in transcriptional activity . Our results thus demonstrate that changes in H3K27me3 abundance are a consequence rather than a cause of transcriptional regulation , and they suggest that oncoprotein-driven changes in gene transcription can alter the pattern of histone modification in cancer cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cancer",
"genetics",
"gene",
"expression",
"genetics",
"epigenetics",
"biology",
"genomics",
"histone",
"modification"
] | 2013 | Ras-Induced Changes in H3K27me3 Occur after Those in Transcriptional Activity |
The identification of MHC class II restricted peptide epitopes is an important goal in immunological research . A number of computational tools have been developed for this purpose , but there is a lack of large-scale systematic evaluation of their performance . Herein , we used a comprehensive dataset consisting of more than 10 , 000 previously unpublished MHC-peptide binding affinities , 29 peptide/MHC crystal structures , and 664 peptides experimentally tested for CD4+ T cell responses to systematically evaluate the performances of publicly available MHC class II binding prediction tools . While in selected instances the best tools were associated with AUC values up to 0 . 86 , in general , class II predictions did not perform as well as historically noted for class I predictions . It appears that the ability of MHC class II molecules to bind variable length peptides , which requires the correct assignment of peptide binding cores , is a critical factor limiting the performance of existing prediction tools . To improve performance , we implemented a consensus prediction approach that combines methods with top performances . We show that this consensus approach achieved best overall performance . Finally , we make the large datasets used publicly available as a benchmark to facilitate further development of MHC class II binding peptide prediction methods .
The activation of CD4+ helper T cells is essential for the development of adaptive immunity against pathogens [1]–[4] . A critical step in CD4+ T cell activation is the recognition of epitopes presented by MHC class II molecules [5] . MHC class II molecules are heterodimers expressed on the surface of professional antigen presenting cells that bind peptide fragments derived from protein antigens [6] . X-ray crystallographic studies demonstrated that the MHC class II epitope binding site consists of a groove and several pockets provided by a β-sheet and two α-helices [7] , [8] . Unlike class I , the class II binding groove is open at both ends . As a result , peptides binding to class II molecules tend to be of variable length , but typically between 13 and 25 residues . A hallmark of the MHC class II binding peptide groove is that there are four major pockets . These pockets accommodate side-chains of residues 1 , 4 , 6 , and 9 of a 9-mer core region of the binding peptide . This core region interaction largely determines binding affinity and specificity [9] . In addition , peptide residues immediately flanking the core region have been indicated to make contact with the MHC molecule outside of the binding groove , and to contribute to MHC-peptide interaction [10] . MHC class II molecules are highly polymorphic , and this polymorphism largely corresponds with differences along the peptide binding groove . However , the binding motifs derived for MHC class II molecules are highly degenerate , and many promiscuous peptides have been identified that can bind multiple MHC class II molecules [11] . Promiscuous peptides are a prime target for vaccine and immunotherapy and computational tools have been developed to facilitate systematic scanning for promiscuous peptides [12] . Computational prediction of MHC class II epitopes is of important theoretical and practical value , as experimental identification is costly and time consuming [13] , [14] . The basis of a successful computational prediction is a sufficiently large set of high quality training data . There are several databases hosting MHC epitope related data such as SYFPEITHI [15] , MHCBN [16] , Antijen [17] , FIMM [18] , HLA Ligand [19] and our own project , the Immune Epitope Database ( IEDB ) [20] , [21] . Information from those databases is , for the most part , extracted from the literature . These databases typically combine data from different sources and different experimental approaches , which can complicate the generation of consistent training and evaluation datasets . The establishment of numerous MHC class II epitope databases has facilitated the development of a large number of algorithms aimed at predicting peptide binding to MHC molecules . Early works focused on finding peptide patterns and deriving motifs for MHC molecules [22]–[24] . With the accumulation of epitope data , more sophisticated algorithms were developed . Several methods have derived scoring matrices that evaluate the contribution to binding of different residues in a peptide based on quantitative binding data ( ARB [25] , SMM-align [26] ) . Others base similar scoring matrices on multiple peptide alignments ( RANKPEP [27] , [28] ) or domain expert knowledge ( SYFPEITHI method [15] ) . By combining the similarities of key residues forming the pockets of the binding groove with quantitative matrices derived from experiments , the TEPITOPE [29] algorithm can predict binding to MHC alleles for which no binding affinities were determined . Other machine learning algorithms that have been applied include hidden Markov models [30] , evolutionary algorithms [31] and linear programming [32] . The MHC class II binding prediction problem has also been modeled with a distance function in a recently developed method PepDist [33] . In addition to the previously listed models that are directly interpretable , “black box” approaches , such as support vector machines [34] and artificial neural networks [35]–[37] , have also been applied to MHC class II binding prediction with success . Despite the large number of available prediction methods , computational prediction of MHC class II epitopes remains a challenging problem . It has been suggested that the prediction performance of class II algorithms is systematically inferior to that of MHC class I epitope prediction methods [25] . To assess the current state of the MHC class II binding predictions , we have here sought to establish a systematic and quantitative benchmark similar to our previous effort for MHC class I molecules [38] . We present a large dataset of unpublished MHC class II-peptide binding affinities that were experimentally determined under uniform conditions . We then proceed to evaluate a set of nine publicly available MHC class II prediction methods using this dataset and systematically compared their performance . Finally , we analyzed the ability of current methods to identify the binding cores of peptides and to predict T-cell responses from peptide sequences .
We assembled a dataset of peptide binding affinities for various MHC class II molecules experimentally measured in our group ( see Materials and Methods for details ) . Table 1 gives an overview of the dataset , encompassing a total of 10 , 017 experimentally determined peptide MHC II binding affinities . These data span a total of 16 human and mouse MHC class II types . The number of unique MHC-peptide affinities measured per type varies greatly , from 3 , 882 for HLA DRB1*0101 , to only 39 for H-2-IEd . Compared to datasets publicly available on the IEDB and other MHC class II epitope databases , our new dataset expands the number of measured peptide-MHC class II interactions significantly for a large number of MHC class II molecules . For example , the number of peptides with known IC50 values for HLA DRB1*0101 was more than tripled with the addition of our new dataset . The MHC class II binding prediction tools evaluated in this study are listed in Table 2 . We included as many prediction methods as possible provided that they ( 1 ) can perform predictions for MHC class II types in our dataset; ( 2 ) were publicly available; and ( 3 ) did not specifically disallow the use of automated prediction retrieval scripts . A total of nine methods matched these criteria . A more detailed description of tested methods is provided in the Materials and Methods section . The binding predictions for peptides in our affinity dataset were extracted from the MHC class II binding prediction tools with custom scripts ( see Materials and Methods for details ) . From the experimental data , peptides were classified into binders ( IC50<1000 nM ) and nonbinders ( IC50≥1000 nM ) based on measured affinities . The performance of the prediction methods was then measured by ROC curves ( see Materials and Methods for details ) . Since the new dataset was never published before , it is equivalent to a blind test . An important exception is the ARB method . Since it was developed at IEDB and was constantly updating with new data , its performance was instead evaluated via 10-fold cross validation . Table 3 shows the prediction performance of the various methods in terms of area under ROC curve ( AUC ) . The ROC curves for tested methods were also plotted in Figure 1 using HLA DRB1*0101 as an example . SVMHC was not evaluated separately , since it implements the same TEPITOPE matrices utilized by PROPRED . When overall performance is assessed by averaging across all available MHC class II molecules SMM-align and PROPRED are associated with the best AUC value ( 0 . 73 ) . The ARB method has the third best performance with an average of AUC of 0 . 71 . When performance on individual MHC class II molecule is examined , the ARB , PROPRED or SMM-align perform best for all but the H-2 IEd molecule , for which RANKPEP gives the best result . Since we restrict our testing to publicly available tools , it is important to point out that the methods were trained on different datasets ( Table 2 ) . Some databases such as MHCPEP only include positive binding data and the lack of nonbinders would be expected to negatively impact some methods that require negative training data . Two of the top performing methods ( SMM-align and ARB ) utilize the IEDB dataset , confirming that the size of the training set maybe an important factor contributing to better performance . PROPRED is among the most accurate MHC class II binding prediction methods , despite being based on the TEPITOPE method developed over eight years ago . The good predictive power of TEPITOPE demonstrates the validity of its approach , based on pocket information derived from MHC class II structures and quantitative peptide binding profiles . The cutoff of 1000 nM to classify peptides into binders and non-binders was chosen following an expert immunologist's recommendation for an immunologically relevant threshold , but it is still somewhat arbitrary . To further our analysis in a systematic fashion , we varied the cutoff from 50 nM to 5000 nM . The changes in cutoffs enable us to evaluate performances of binding prediction to identify peptides with different affinities . A cutoff of 50 nM focuses on identifying strong binders , while a cutoff of 5000 nM will identify all including very weak binders . The results of the evaluation using different cutoffs are shown in Dataset S1 . For MHC molecules with large number of binding affinities ( such as HLA DRB1*0101 ) , varying the cutoff has little impact on the AUC values . For datasets with smaller number of binding affinities ( such as H-2-IEd ) , the change in AUC values is more significant . Despite the variations in AUC values introduced by different cutoffs , the relative performance of different methods remains largely the same , suggesting our conclusion of different methods' performance is not strongly dependent on the cutoff used to decide binders . A key difference between MHC class I and class II molecules is that the binding groove of class II molecules is open at both ends [7] , [8] . As a result , the length of peptides binding class II molecules can vary considerably , and typically range between 13 and 25 amino acids long . Thus , a requisite for all MHC class II binding prediction approaches is the capacity to identify within longer sequences the correct 9-mer core residues that mediate the binding interaction [9] . All methods included in our study explicitly predict cores when they predict MHC class II binding peptides . They either predict binders as 9-mer peptides , or clearly state in prediction the location of the predicted 9-mer core . We next analyzed whether the various class II prediction tools can accurately identify the 9-mer cores of a binding peptide . We extracted MHC-peptide complex structures from the Research Collaboratory for Structural Bioinformatics ( RCSB ) Protein Data Bank ( PDB ) . A total of 29 structures associated with 14 different MHC class II molecules were identified ( Table 4 ) . For each method we compared the predicted cores with the true cores extracted from crystal structures . The results are shown in Table 5 . The PROPRED method based on TEPITOPE was associated with the best performance , with all the predicted cores matching the cores determined by PDB structures . This is in good agreement with the fact that TEPITOPE is directly based on experimental assays . SYFPEITHI is the second best method with an accuracy of 0 . 9 . However , this result should be interpreted with caution since seven of the nine correctly predicted peptides are documented in the SYFPEITHI database . Apart from PROPRED and SYFPEITHI , the method most effective in predicting binding affinity ( SMM-align ) is also the method with highest accuracy in predicting 9-mers cores , with an accuracy of 0 . 625 . RANKPEP and SVRMHC come next with accuracies about 0 . 55 . The remaining three methods had limited success ( 21–25% ) , although they still perform above random prediction ( the probability to randomly guess the right core for a 15-mer peptide is 1 out of 7 or 0 . 143 ) . Overall , these data suggest that correctly aligned cores contribute to the superior performance of PROPRED and SMM-align , and that there is substantial room to improve the quality of the core predictions of other methods . The ultimate goal of MHC binding peptide prediction is to identify epitopes that activate T cells . Recognition of a peptide bound to an MHC molecule by a T cell receptors is the critical step in this activation , and binding of peptide to the MHC molecule is obviously a necessary requirement [39] . In a separate study carried out in our lab , a set of 664 peptides overlapping the LCMV proteome were tested for their abilities to promote H-2 IAb specific IFN gamma production from CD4+ T cells in splenocytes from previously LCMV infected mice ( manuscript in preparation ) . These peptides provided an ideal test set to evaluate MHC class II binding predictions as a tool to identify peptides that trigger an immune response . For each of the 664 peptides , we obtained H-2 IAb binding predictions from the five methods in our study that cover H-2 IAb following exactly the same procedures as predictions of simple binding . We then evaluated the methods' performance in predicting which peptides triggered an immune response . The ROC curves quantifying the performance of each method are shown in Figure 2 . The Consensus method is the best performing methods with AUC of 0 . 89±0 . 05 . ARB is the second best performing method with an AUC of 0 . 85±0 . 05 . SMM-align and RANKPEP have similar performance with AUC about 0 . 76±0 . 08 and 0 . 78±0 . 07 , respectively . MHCPRED and MHC2PRED do not perform as well , with AUC values of about 0 . 67±0 . 12 and 0 . 36±0 . 1 ( standard deviations calculated by bootstrapping with replacement ) . Except MHCPRED , every other method's performance in this evaluation compared favourably to that in predicting peptide binding . Overall , the ranking of prediction performances is well in concert with that for predicting peptide binding , specifically when taking into account the high standard deviations of AUC values . These large standard deviations are due to the limited number of positive datapoints in the set utilized . To further analyze the performance of the T cell activation prediction , we classified peptides into predicted binders and non-binders . Since different methods produce scores on different scales , we adopt a rank based classification in that we classify the top 10% highest scoring peptides as binders . We then calculated sensitivity and positive predictive value ( PPV ) for each method ( Table 6 ) . These two measures were chosen since we are primarily interested in identifying T cell activating peptides while minimizing the number of false positive predictions . In our system , sensitivity is the percentage of peptides activating T cells predicted to be binders and PPV is the percentage of predicted binders that actually activate T cells . The data in Table 6 show that results of these two measures are largely consistent with the AUC results . Methods with high AUCs tend to have high PPV and sensitivity . Only the consensus method has sensitivity above 50% , indicating that 5 out of 6 methods missed more than half of the T cell activating peptides when top 10% ranked peptides are classified as binders . In addition , the consensus method also had the highest PPV value of 9 . 4% , making it again the best prediction method by this measure . The overall low PPV values are expected , as many peptides that are capable of binding MHC are not recognized by T cells following a natural infection , due to other factors such as peptide processing and the available T cell repertoire . Our evaluation of prediction performance suggests that in all cases there is clearly room for improvement , and that no single method is dominantly better than all others . Motivated by the success of a consensus prediction approach to map MHC class I epitopes in vaccinia virus [40] , we implemented the same approach for MHC class II binding predictions . This consensus approach is based on calculating the median rank of the top three predictive methods for each MHC class II molecule ( see materials and methods for details ) . The consensus prediction performance is shown in the last column of Table 3 for the 14 MHC alleles for which three or more predictions were available . For ten of these fourteen alleles , the consensus method gives similar or higher performance than the best individual method . For each of the remaining four datasets , a single prediction performs better ( i . e . , ARB for DRB1*1302 , SMM-align for DRB1*1101 and DRB1*1501 , and PROPRED for DRB1*0405 ) . In terms of overall performance across all molecules in our dataset , the consensus method outperforms all individual MHC class II prediction methods . The MHC-peptide affinity , MHC-peptide structure and T cell activation datasets are available as supplemental material at http://mhcbindingpredictions . immuneepitope . org/MHCII . These data are presented in this paper for immediate access by the immunology and bioinformatics community . We are currently in the process of depositing these data into the IEDB , making them available through the epitope informatics framework of the IEDB .
In this study we have presented a comprehensive dataset for the systematic evaluation of MHC class II peptide binding prediction methods . This dataset consists of three components . The first component is a large set of 10 , 017 quantitative peptide-binding affinities for 16 MHC class II types that significantly expands the amount of publicly available data . These data were generated under identical experimental conditions and comprise affinities for binders as well as non-binders . The second component is a set of non-redundant structures of MHC class II molecules complexed with peptide ligands compiled from the PDB . This set of structures provided a “gold standard” for evaluating the ability of prediction methods to locate the 9-mer core of epitopes . The last component is a set of 664 peptides that has been tested experimentally to determine their ability to stimulate CD4+ T cells from widely utilized C57BL/6 ( H-2b ) strain of laboratory mice . Together , these datasets serve as a benchmark set to facilitate the development and testing of algorithms for predicting peptide binding to MHC as well as T-cell responses . Several previous studies have compared the performances of various MHC class II binding prediction methods [41]–[43] . The Borras-Cuesta study [43] from 2000 only had a limited number of peptides , alleles and methods to compare . The two recent studies were published after we finished our initial comparison . Gowthaman et al [42] compared six commonly used method with data spanning seven MHC class II alleles . However , their evaluation dataset comprised only 179 peptides , limiting the significance of their results . Rajapakse et al [41] compared their multi-objective evolutionary algorithms ( MOEA ) with five other algorithm using two datasets . The first dataset consisted of 1 training and 10 testing datasets on HLA DRB1*0401 assembled from different sources . The second dataset was extracted from the IEDB and comprised more than 5 , 000 peptides covering 16 MHC class II alleles . We couldn't include MOEA in our comparison since it is not publicly available at the moment . Despite the difference in datasets used in comparison , their conclusion is consistent with ours in that SMM-align , TEPITOPE and ARB are the better performing methods . We have carried out a comprehensive unbiased evaluation of existing MHC class II epitope prediction algorithms using these datasets . Except binding prediction for ARB , all the other MHC class II prediction algorithms are evaluated in a completely blinded fashion . In our analysis , the better performing methods proved to be those that are based on quantitative matrices extended by method specific features . For example , SMM-align is the only method tested that considers the contribution of residues outside of the binding groove , and TEPITOPE is the only method whose matrices are based on experiments aimed to determine individual amino acid's contribution to binding . Merely using quantitative matrices alone is not sufficient to ensure good performance , since pure position specific scoring matrix based methods such as RANKPEP and SYFPEITHI do not perform as well . One potential reason for the differential performance of various methods is the likely different number of data points utilized by the various methods in the training stage . In this respect , we anticipate that the datasets described herein , and now made publicly available , could be utilized to retrain several of the methods and further increase their performance . Despite the large number of existing MHC class II epitope prediction methods , the best performance is generally not as good as that for MHC class I methods . Indeed , it is notable that the majority of methods examined in the present study have also been employed to make predictions for MHC class I peptide binding , and almost invariably their performance is appreciably better in the context of class I [38] . For example , when SMM [44] was applied to predict epitopes for several MHC class I molecules , it achieved an average AUC of 0 . 874 , which is substantially higher than that for class II ( 0 . 783 ) . In an attempt to identify what limits the performance of MHC class II binding prediction , we tested the ability of prediction methods to identify the 9-mer peptide cores revealed in crystal structures of MHC-peptide complexes . Except for PROPRED and SYFPEITHI , the methods examined performed poorly , suggesting that difficulties in identification the correct binding core contribute to the inferior performance of class II binding prediction . It is noteworthy that the two methods with the best core predictions do not take all positions of a peptide into account when making binding predictions , but rather focus on anchor positions in the peptide . This may explain why especially the ARB method performs much poorer in the core identification rather than the binding predictions: It treats all positions in the peptide identically and relies on automated peptide alignments to derive an overall peptide profile . While this inclusion of weakly interacting positions can be an advantage to predict overall peptide binding , it may lower the accuracy when picking the correct core . In an attempt to improve upon the prediction performance realized by individual prediction tools , we implemented a consensus approach for class II binding predictions . The consensus approach was found to clearly outperform each individual prediction approach when measured over the entire dataset , and provided the best predictions for 10 out of 14 molecules . This shows that the consensus approach is just as useful for MHC class II peptide binding prediction as its recent successful application for MHC class I molecules [40] . In a smaller study addressing 3 different prediction methods in the context of a single DR type , Mallios previously came to a similar conclusion [45] . Other types of meta approaches have been successfully applied to MHC binding prediction . For example , Mallios [46] has used an iterative stepwise discriminant analysis meta-algorithm to successfully classify binders and non-binders for HLA-DR1 . Stern and co-workers effectively used a two-dimensional dot plot to combine the prediction results of SYFPEITHI and TEPITOPE [47] . Finally , Trost et al [48] have reported achieving greater accuracy in MHC class I binding predictions by combing results from multiple prediction tools . Compared to these methods , our median rank approach does not depend on the absolute values of scores and it has exceptional scalability since typical sorting algorithms have running times proportional to n log n where n is the number of cases needed to be sorted . Overall , it is astonishing that the systematic use of consensus predictions comes rather late ( see Mallios [45] , [46] ) to the problem of MHC peptide binding since consensus approaches have for quite some time proven their superiority in a number of fields , notably protein structure prediction [49] . In any case , it is also likely that the remarkable increase in performance obtained by the use of the consensus approach hinges on the fact that it combines information derived from methods trained on large numbers of data points with methods incorporating structural considerations leading to effective core predictions . We are currently working on development of algorithms specifically combining these two different features . We also tested the ability of MHC class II binding prediction methods to predict a peptide's ability to activate CD4+ T cells . Most of the methods were associated with good performance . This was somewhat surprising since T cell activation is a multi-step process where multiple signals are needed for successful activation [50]–[52] . In addition , a peptide that binds well to MHC molecules is not necessarily a good stimulator for T-cell response as different amino acids are interacting with T cell receptor . It is important to point out that the performance was based on a set of 664 peptides of which only 9 activated CD4+ T cells . The limited number of positive cases makes the ROC curve jagged and the AUC values calculated less robust . Despite the encouraging AUC values achieved by several methods , it is still necessary to test a large number of peptides to identify most of the T cell activating peptides . In addition , all those methods still have high numbers of false positives peptides that are predicted binders but will not activate T cells . Since experimental efforts to test T cell activation are even more time consuming than testing peptide-MHC binding , significant efforts are needed to develop tools that can identify T cell activating peptides with high sensitivity and specificity . In conclusion , we have presented a set of benchmarks to facilitate the evaluation and development of MHC class II binding predictions . While several good methods are available , these do not reach the performance of those for MHC class I molecules . We have shown that a simple and robust consensus approach can improve the prediction performance for the great majority of the MHC class II molecules tested . Finally , we speculate that novel approaches that capture distinct features of MHC class II peptide interactions could lead to more successful predictions than the current approaches , which are commonly developed as extensions of MHC class I predictions .
Peptides utilized for the assessment of MHC binding , antigenicity and immunogenicity were purchased as crude material from Mimotopes ( Minneapolis , MN and Clayton , Victoria , Australia ) , Pepscan Systems B . V . ( Leylstad , Netherlands ) or A and A Labs ( San Diego , CA ) . Quality control analyses of crude syntheses were performed by mass spectrometry on randomly selected peptides . Peptides selected for additional deconvolution and HLA peptide binding assays were resynthesized by A and A as purified material . Peptides were purified to >95% by reversed-phase HPLC , and the purity assessed by amino acid sequence and/or composition analysis . Quantitative assays to measure the binding affinities of peptides to purified soluble class II molecules are based on the inhibition of binding of a radiolabeled standard peptide . Binding assays were performed essentially as described previously [13] , [53] . Briefly , 0 . 1–1 nM radiolabeled peptide was coincubated for 2 days at room temperature with 1 µM to 1 nM purified MHC in the presence of a cocktail of protease inhibitors . Following a two-day incubation , the amount of MHC bound labelled peptide was determined by capturing MHC/peptide complexes on LB3 . 1 antibody coated Lumitrac 600 microplates ( Greiner Bio-one , Longwood , FL ) , and measuring bound cpm using the TopCount microscintillation counter ( Packard Instrument Co . , Meriden , CT ) . Individual peptides were typically tested in 3 or more independent experiments for its capacity to inhibit the binding of the radiolabeled peptide . The concentration of peptide yielding 50% inhibition of the binding of the radiolabeled peptide was calculated . Under the conditions used , in which [label]<[MHC] and IC50≥[MHC] , the measured IC50 values are reasonable approximations of the true Kd values . The binding affinities are expressed in terms of IC50 and are capped at 50 , 000 nM , reflecting the experimental sensitivity threshold . The assembled MHC class II peptide binding affinities are listed in Table 1 . The peptide binding affinities for various MHC class II molecules were generated in the context of various projects currently ongoing in our laboratory . Because they have been recently generated , to the best of our knowledge , none of the binding affinities in this dataset has been previously published . This assessment was confirmed by comparing our dataset to publicly available records contained in the IEDB ( Table 1 ) or elsewhere . There are total 10 , 017 measured affinities in our dataset spanning thirteen human and three mouse MHC class II types . Peptides for 114 proteins from 30 organisms were synthesized and tested . While peptide sizes ranged form 9 to 37 amino acids , the vast majority of the measured affinities are for 15-mers ( 9 , 632 out of 10 , 017 ) . The present dataset is currently in the process of being deposited in the IEDB . Structures of MHC class II were retrieved from the Protein Data Bank with a keyword search ( using keyword “MHC class II” ) . The retrieved structures were then examined to select complexes have epitopes with at least 9 amino acids . In addition , the structures were examined to identify entries with identical MHC and binding peptide sequences . For duplicated structures of the same MHC and epitope , we retained the structure with the highest resolution . The final dataset contains 29 non-redundant structures . The eight MHC class II binding prediction tools evaluated in this study are listed in Table 2 . Five of the prediction methods are based on various scoring matrices . The method developed at IEDB utilizes the Average Relative Binding ( ARB ) matrix [25] . PROPRED [54] and SVMHC [55] are web servers based on TEPITOPE's pocket profile [29] . Both SYFPEITHI [15] and RANKPEP [28] use position specific matrices . Another matrix based approach , SMM-align [26] , utilizes the stabilized matrix method ( SMM [44] ) , but introduces a novel step to identify peptide binding cores , which makes it applicable to MHC class II predictions . Two of the methods , SVRMHC [56] and MHC2PRED ( http://www . imtech . res . in/raghava/mhc2pred/index . html ) , apply support vector machine or support vector regression to predict epitopes . Finally , MHCPRED is a quantitative structure activity relationship ( QSAR ) regression based method [57] . Three of the nine methods , ARB , MHC2PRED and SMM-align , give predictions in terms of the quantitative affinity of a peptide for a MHC class II molecule . The predictions of the other six methods are given as a score which is not directly translatable into an affinity of peptide-MHC binding . In terms of the number of MHC class II types covered , the two TEPITOPE based methods ( PROPRED and SVMHC ) have the broadest coverage with 51 types , 11 of which also appear in our dataset . The next most comprehensive method is RANKPEP which covers 46 types , 16 of which overlap with our dataset . ARB , MHC2PRED and SMM-align make predictions for about 20 MHC class II types and the majority of the types ( 15 to 16 ) also appear in our dataset . The three remaining methods ( MHCPRED , SVRMHC and SYFPEITHI ) have less coverage , as they only predict peptide binding for 5 to 6 MHC class II types in our dataset . Table 2 also lists the dataset used by each method to train their predictive models . Training on larger sets of data would be expected to yield better performance when tested on independent new data . In this context , the IEDB has HLA-DRB1*0101 binding information for 1390 peptides , AntiJen for 730 , and MHCBN for 588 . By contrast , SYFPEITHI lists only 42 entries for HLA-DRB1*0101 . Thus the ARB and SMM-align methods which use data from the IEDB , had access to the largest training set compared to other methods , while the SYFPEITHI method had access to the smallest dataset . We identified eight publicly available MHC class II prediction tools through literature search and the IMGT link list at http://imgt . cines . fr/textes/IMGTbloc-notes/ . For each tool , we mapped the MHC types for which predictions could be made to the four-digit HLA nomenclature ( e . g . , HLA-DRB1*0101 ) . If this mapping could not be done exactly , we left that type/tool combination out of the evaluation . For example , HLA-DR4 could refer to HLA-DRB1*0401 , DRB1*0402 etc , which do have distinct binding specificities . For the ARB evaluation , the 10-fold cross validation results stored at IEDB was used to estimate performance since ARB was trained on datasets overlapping with the one used in this study . For the other seven tools in the evaluation , we wrote python script wrappers to automate prediction retrieval . For the SYFPEITHI prediction , we patched each testing peptide with three Glycine residues at both ends before we submitted it for prediction . This was recommended by the creators of SYFPEITHI method to ensure that all potential binders are presented to the prediction algorithm . For all other methods , the original testing peptides were submitted directly for prediction . Peptide sequences were sent to the web servers one at a time and predictions were extracted from the server's response . To assign a single prediction for peptides longer than nine amino acids in the context of tools predicting the affinity of 9-mer core binding regions , we took the highest affinity prediction of all possible 9-mers within the longer peptide as the prediction result . For each MHC class II molecules whose binding can be predicted by three or more algorithms , we employed the following approach to generate a consensus prediction . First , we selected the top three methods that give the best performance . For each method , the tested peptides are ranked by their scores with higher ranks for better binders . For each tested peptide , the three ranks from different methods are then taken and the median of the three is calculated . This median rank is taken as the consensus score . Receiver operating characteristic ( ROC ) curves [58] were used to measure the performance of MHC class II binding prediction tools . For binding assays , the peptides were classified into binders ( experimental IC50<1000 nM ) and nonbinders ( experimental IC50≥1000 nM ) , which was determined as a practical cutoff in a previous study [59] . For CD4+ T cell activation assays , the peptides were classified into T-cell epitopes ( experimental SFC count≥100 ) or non-epitopes ( experimental SFC count <100 ) . For a given prediction method and a given cutoff for the predicted scores , the rate of true positive and false positive predictions can be calculated . An ROC curve is generated by varying the cutoff from the highest to the lowest predicted scores , and plotting the true positive rate against the false positive rate at each cutoff . The area under ROC curve is a measure of prediction algorithm performance where 0 . 5 is random prediction and 1 . 0 is perfect prediction . The plotting of ROC curve and calculation of AUC are all carried out with the ROCR [60] package for R [61] . C57BL/6 ( H-2b ) mice were purchased from The Jackson Laboratory ( Bar Harbor , ME ) , and infected intraperitoneally with 2×105 PFU of LCMV Armstrong ( i . p . ) . Spleens were harvested eight days post infection , and IFN-γ ELISPOT assays were performed as previously described [62] using CD4+ T cells isolated with anti-CD4+ magnetic beads ( Miltenyi Biotech Inc . , Auburn , CA ) . Experimental values were expressed as the mean net spots per million CD4+ cells ±SD for each peptide pool or individual peptide . For the initial screening of the 83 pools , responses against each pool were considered positive if a ) the number of spot forming cells ( SFCs ) /106 CD4+ T cells exceeded the absolute value of the mean negative control wells ( effectors plus APCs without peptide ) by two-fold , b ) the value exceeded 200 SFCs/106 CD4+ cells and c ) these conditions were met in at least two replicate independent experiments . Positive pools were deconvoluted into their eight individual components and tested again , to determine which individual peptides were responsible for the pooled IFN-γ response . Responses against individual peptides were considered positive if they exceeded the threshold of the mean negative control wells ( effectors plus APCs without peptide ) by at least 2 standard deviations and exceeded a threshold of 200 SFCs/106 CD4+ cells . | A critical step in developing immune response against pathogens is the recognition of antigenic peptides presented by MHC class II molecules . Since experiments for MHC class II binding peptide identification are expensive and time consuming , computational tools have been developed as fast alternatives but with inferior performance . Here , we carried out a large-scale systematic evaluation of existing prediction tools with the aim of establishing a benchmark for performance comparison and to identify directions that can further improve prediction performance . We provide an unbiased ranking of the performance of publicly available MHC class II prediction tools and demonstrate that the MHC class II prediction tools did not perform as well as the MHC class I tools . In addition , we show that the size of training data and the correct identification of the binding core are the two factors limiting the performance of existing tools . Finally , we make available to the immunology community a large dataset to facilitate the evaluation and development of MHC class II binding prediction tools . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology",
"computational",
"biology"
] | 2008 | A Systematic Assessment of MHC Class II Peptide Binding Predictions and Evaluation of a Consensus Approach |
HIV-infected individuals have deficient responses to Yellow Fever vaccine ( YFV ) and may be at higher risk for adverse events ( AE ) . Chronic immune activation–characterized by low CD4/CD8 ratio or high indoleamine 2 , 3-dioxygenase-1 ( IDO ) activity—may influence vaccine response in this population . We prospectively assessed AE , viremia by the YFV virus and YF-specific neutralizing antibodies ( NAb ) in HIV-infected ( CD4>350 ) and -uninfected adults through 1 year after vaccination . The effect of HIV status on initial antibody response to YFV was measured during the first 3 months following vaccination , while the effect on persistence of antibody response was measured one year following vaccination . We explored CD4/CD8 ratio , IDO activity ( plasma kynurenine/tryptophan [KT] ratio ) and viremia by Human Pegivirus as potential predictors of NAb response to YFV among HIV-infected participants with linear mixed models . 12 HIV-infected and 45-uninfected participants were included in the final analysis . HIV was not significantly associated with AE , YFV viremia or NAb titers through the first 3 months following vaccination . However , HIV–infected participants had 0 . 32 times the NAb titers observed for HIV-uninfected participants at 1 year following YFV ( 95% CI 0 . 13 to 0 . 83 , p = 0 . 021 ) , independent of sex , age and prior vaccination . In HIV-infected participants , each 10% increase in CD4/CD8 ratio predicted a mean 21% higher post-baseline YFV Nab titer ( p = 0 . 024 ) . Similarly , each 10% increase in KT ratio predicted a mean 21% lower post-baseline YFV Nab titer ( p = 0 . 009 ) . Viremia by Human Pegivirus was not significantly associated with NAb titers . HIV infection appears to decrease the durability of NAb responses to YFV , an effect that may be predicted by lower CD4/CD8 ratio or higher KT ratio .
Effective antiretroviral treatment ( ART ) drastically improved clinical outcomes for people living with HIV . However , these patients still present increased risk of death , higher prevalence of comorbidities , and impaired responses to vaccines [1–6] . Prior studies have shown impaired Yellow Fever vaccine ( YFV ) immunogenicity among HIV-infected persons is associated with detectable HIV viral load ( VL ) [7–12] and lower CD4 T cell counts [11] . However , it is still unclear whether reduced YFV antibody response among HIV-infected individuals is caused by a blunted initial response , decreased persistence of antibodies , or both . Moreover , predictors of YFV immunogenicity among patients with effective and early ART are not well known . More recently , studies including patients with early initiation of ART have suggested a negative effect of persistent immune activation on responses to Influenza vaccine [13 , 14] , Neisseria meningitis vaccine [15] and YFV [16 , 17] in both HIV-infected and–uninfected individuals . This is consistent with previous studies that demonstrate excessive immune activation and inflammation predict residual morbidity and mortality in treated HIV-infected patients [18–20] . A range of biomarkers has been used in different settings to quantify persistent immune activation [20] . One increasingly appraised indirect biomarker is the ratio of CD4 to CD8 T lymphocytes , or CD4/CD8 ratio . Previous studies have shown that CD4/CD8 ratio correlates with markers of CD8 T cell activation , and a lower CD4/CD8 ratio predicts higher risk of non-Aids events and mortality among ART-treated HIV-infected patients [21–23] . Furthermore , a low CD4/CD8 ratio is also strongly associated with the activity of Indoleamine 2 , 3-dioxygenase-1 ( IDO ) , an enzyme expressed by activated myeloid cells in HIV and other inflammatory conditions that causes adaptive immune defects . IDO catabolizes tryptophan ( T ) to kynurenine ( K ) and other metabolites that may contribute to proliferative lymphocyte defects , regulatory T cell expansion , microbial translocation and immune activation in treated HIV infection [24] . Therefore , elevated IDO activity ( as measured by plasma KT ratio ) may also indicate adaptive immune dysfunction in this population . Finally , chronic co-infection with Human Pegivirus has been associated with reduced innate and adaptive immune activation among HIV-infected patients in prior studies [25–27] . An additional relevance of YFV in HIV-infected patients concerns the theoretic higher risk of YFV-associated severe adverse events ( AE ) in this population [28 , 29] . YFV is produced from the 17D or 17DD attenuated viral strains , and although mechanisms for YFV-associated AE are not completely elucidated , immunosuppressed patients and persons at extremes of age are considered at increased risk [28 , 30] . It is hypothesized that immune response fails to contain the vaccine virus replication , typically seen only in the first 3–5 days after vaccination , leading to uncontrolled viral dissemination and clinical disease [31] . In this study , HIV-infected patients and controls referred to receive YFV were enrolled and prospectively followed for 1 year after vaccination . We addressed clinical and laboratory AE and viremia by the YFV virus . In addition , we investigated if HIV status was associated with titers of Yellow Fever ( YF ) -specific neutralizing antibodies ( NAb ) in the first 3 months following vaccination ( henceforth defined initial YFV immunogenicity ) and one year following vaccination ( henceforth defined persistence of YFV immunogenicity ) . Finally , we explored if correlates of immune activation including CD4/CD8 ratio , KT ratio , and Human Pegivirus co-infection predict YFV response among HIV-infected subjects .
Subjects aged 18 years old and older who were referred to receive YFV at Clinics Hospital in Sao Paulo , Brazil , between October 2011 and April 2014 were screened for participation . Potential participants were evaluated by an attending physician who determined whether YFV was indicated based on risk of exposure to wild YF and YFV contraindications as defined by National Guidelines . The Guidelines do not recommend YFV to pregnant and breastfeeding women , subjects under immunosuppressive medications and patients with conditions such as cancer and thymus dysfunction . HIV-infected patients with a CD4 T cell count above 350/ml measured in the previous 4 months were considered eligible for vaccination . At enrollment , HIV-negative persons underwent a rapid HIV-test . For both groups , participants with immunosuppressive conditions other than HIV infection were excluded . These included diabetes , chronic liver or kidney diseases , any type of cancer ( except resolved Kaposi Sarcoma ) , and use of systemic immunosuppressive therapy in the last 3 months . Female participants in reproductive age underwent a pregnancy test at enrollment . At enrollment , medical history and date of previous YFV was obtained , if applicable . A blood sample was collected for assessment of baseline complete blood count and liver enzymes , CD4 and CD8 T cell counts , plasma KT ratio and Human Pegivirus viremia . HIV-infected participants had HIV VL measured at baseline and all subsequent visits . Participants were followed on days 3 , 5 , 7 , 14 , 28 , 56 , 84 , and 365 after vaccination . We measured viremia by the YFV virus on days 3 , 5 , 7 , and 14 after vaccination , and measured titers of YF-specific NAb at baseline and on days 7 , 14 , 28 , 56 , 84 and 365 after vaccination . We also collected data on spontaneous and solicited clinical AE , as well as laboratory AE on visits 3 , 5 , 7 , 14 and 28 after vaccination . We measured CD4 , CD8 T cells and CD4/CD8 ratio in all visits ( Fig 1 ) . Clinical and laboratory AE were assessed as binary variables , defined as positive if the participant had any clinical AE , or any clinically significant laboratory AE at visits 3 , 5 , 7 , 14 or 28 , and negative otherwise . Laboratory AE were considered clinically significant if graded ≥2 according to the National Institutes of Allergy and Infectious Diseases’ Division of AIDS Table for Grading the Severity of Adult and Pediatric AE [32] . Viremia by YFV virus was assessed both as numeric and binary variable . The binary variable for YFV viremia was defined as positive if the participant had a detectable measurement ( >200 copies/ml ) on days 3 , 5 , 7 or 14 after vaccination , and negative otherwise . The HIV VL was determined by reverse-transcriptase ( RT ) -PCR using Amplicor HIV-1 Monitor Test ( Roche Diagnostic Systems , NJ , USA ) , with a lower detection limit of 40/mm3 . CD4 and CD8 T cell counts were determined by flow cytometry ( FACSCalibur , BD Biosciences , CA , USA ) using Multitest reagent ( BD Biosciences ) . Kynurenine and tryptophan were quantified on cryopreserved plasma samples by liquid chromatography–tandem mass spectrometry as previously described [33] . Human Pegivirus RNA was extracted from 140μl serum samples using QIAamp Viral RNA Mini Kit ( QIAGEN Inc . , California , USA ) , according to manufacturer’s instructions . A 5μl aliquot of extracted RNA was used to perform qRT-PCR with SuperScript III Platinum One-Step Quantitative RT-PCR System with ROX kit ( Life Technologies ) , with primers and a TaqMan probe that amplified and quantified a fragment of 72-bp of the 5' untranslated region ( 5'UTR ) . The reaction was made with 0 . 5μl of SuperScript III RT/Platin Taq Mix , 12 . 5μl of 2X Reaction Mix with ROX , 0 . 75μl of 10μM Forward primer RTG1 ( 5’GTGGTGGATGGGTGATGACA3’ ) ( Sigma ) , 1 . 25μl of 10μM Reverse primer RTG2 ( 5’GACCCACCTATAGTGGCTACCA3’ ) ( Sigma ) , 0 . 4μl of 25 μM TaqMan probe ( [6’FAM]CCGGGATTTACGACCTACC [TAMRA-6-FAM] ) ( Life Technologies ) , and reaction final volume of 25μl was completed with DEPC-treated water . cDNA synthesis was performed during the first 15 minutes at 50°C . After 2 minutes at 95°C , amplification and quantification were performed during 40 cycles with the following times and temperatures: 95°C , 15 seconds; 60°C , 30 seconds . The reading of FAM fluorescence was made during annealing period at 60°C . For measurement of YFV viremia , total RNA was extracted from 140μl of plasma using QIAamp RNA Blood Mini Kit ( Qiagen , Hilden , Germany ) and eluted in 60μl of elution buffer . cDNA was obtained through RT reaction using 10μl of extracted RNA , 300ng of random primer ( Amersham Biosciences , Piscataway , NJ , USA ) ; 10U/μl of Super Script II RT ( Invitrogen , Carlsbad , CA , USA ) in a buffer solution with 0 . 25U/μl of ribonuclease inhibitor ( Invitrogen ) and 0 . 5mM deoxyribonucleotide triphosphates ( Invitrogen ) , at final volume of 20μl . The reaction was incubated at 45°C for 90 minutes . Five μl of cDNA was added to 20μl of TaqMan Master Mix ( Applied Biosystems , Foster City , CA , USA ) and amplified by RT-PCR using the following primers and probe: ( YF-NS5_F ) 5’-GCACGG ATGTAACAGACTGAAGA-3’; ( YF-NS5_R ) 5’-CCAGGCCGAACCTGTC AT-3’ and ( YF-NS5Probe ) 5’-FAM-CGACTGTGTGGTCCGGCCCATC-3’–TAMRA [34] . The product was amplified using optical detection system layout of BioRad ICycler for 45 cycles at the following settings: 10 min , 95°C; 45 cycles of 15s for 94°C and 60s for 60°C . NAb titers against YF virus were measured by Plaque Reduction Neutralization Test ( PRNT ) performed at Virologic Technology Laboratory of Bio-Manguinhos ( LATEV , FIOCRUZ , Rio de Janeiro ) as previously described [16] . The study was approved by the Ethics Committee at Clinics Hospital in University of Sao Paulo . Upon participation , all participants signed an informed consent form . HIV tests were performed with pre and post-test counseling , and all individual identifiable information was maintained in secured cabinets and electronic files . Groups were compared using Wilcoxon rank-sum test for continuous variables and Fisher’s exact test for categorical variables . Titers of YF-specific NAb , CD4 and CD8 T cell counts were log-transformed to approximate normal distribution , and antilog transformation was required for model interpretation . The effect of HIV status on levels of YFV viremia , and on initial YFV immunogenicity were investigated using mixed models with robust standard errors , adjusted for age , sex , previous YFV and interaction between HIV and visits . The effect of HIV status on persistence of YFV immunogenicity was investigated using a linear regression model adjusted for age , sex , previous YFV and baseline values of YF-specific NAb titers . The effects of T CD4 and T CD8 cell count , detectable HIV VL , CD4/CD8 ratio , KT ratio , and Human Pegivirus viremia on YF-specific NAb titers among HIV-infected patients were investigated using mixed models adjusted for age , baseline NAb titers and HIV VL . Correlations between NAb titers and CD4/CD8 ratio or KT ratio were explored using Spearman rank correlation . For all analysis , we assumed a two-sided alpha error of 0 . 05 . All analyses were performed in Stata version 13 . 1 ( StataCorp . College Station , TX: StataCorp LP ) . We calculated sample size based on the impact of HIV status on titers of YF-specific NAb in the first 3 months following vaccination , using estimates of effect size and standard deviation from a prior study published by our group [16] . Since the analysis plan included mixed models for repeated outcomes , we assumed a 20% reduction in error variance and estimated a final sample of 33 participants per group using conventional means comparison .
Any clinical AE was reported by 6 ( 50% ) participants in the HIV-infected group , and 22 ( 48 . 9% ) controls . All reported clinical AE ( local pain , tenderness and redness; nausea , myalgia , fatigue , dizziness and fever ) were mild and self-limited , as were laboratory AE–anemia , neutropenia , lymphopenia , thrombocytopenia and liver enzymes elevation–which were detected in 3 ( 25% ) individuals in the HIV-infected group , and 13 ( 28 . 9% ) controls . Viremia by the YFV virus was detected in at least one visit in 40% of HIV-infected participants and 34% of controls . Maximum detected viremia was 11210 copies/mL in one HIV-uninfected participant at day 5 after vaccination; in the HIV-infected group , highest measured viremia was also observed at day 5 ( 4197 copies/mL ) . HIV status was not statistically associated with levels of viremia by the YFV virus ( p-value = 0 . 99 for the overall effect of HIV on YFV viremia on days 3 , 5 , 7 and 14 ) . At baseline , 4 HIV-infected participants ( 33% ) and 17 controls ( 38% ) had levels of NAb considered seropositive for a cutoff of 794 mUI/mL as defined by the referent laboratory [35] . If only participants with a previous YFV were considered , 3 HIV-infected ( 75% ) and 15 controls ( 94% ) had seropositive NAb titers . At 28 days after vaccination , all participants in both groups were seropositive for YF , and at one year following vaccination , 11 HIV-infected participants ( 92% ) and 43 controls ( 96% ) maintained seropositive YF-specific NAb . We failed to find a statistically significant difference between groups defined by HIV status on initial YFV immunogenicity in either visit individually or overall in a mixed model adjusted for age , sex and previous YFV ( Table 2 ) . The model predicted lower YF-specific NAb titers for women; in average , women had 0 . 33 times the titers observed for men ( 95% CI 0 . 17–0 . 66 , p = 0 . 002 ) . As expected , compared to individuals without previous YFV , those who reported a previous YFV had higher NAb titers ( fold change 13 . 69 , 95% CI 7 . 12–26 . 30 , p<0 . 001 ) . Persistence of YFV immunogenicity was significantly lower in HIV-infected participants compared to controls in a mixed model adjusted for age , sex , previous YFV and baseline NAb titers . In average , HIV-infected individuals had 0 . 32 times the NAb titers observed for HIV-uninfected participants at one year after vaccination ( 95% CI 0 . 13–0 . 83 , p = 0 . 021 ) . We found no statistically significant effect of age , sex or previous YFV on persistence of NAb ( Table 3 ) . In the exploratory analysis restricted to HIV-infected patients , higher CD4/CD8 ratio and lower KT ratio predicted higher YF-specific NAb titers; in average , for each 10% increase in CD4/CD8 ratio , post-baseline NAb titers were 21% higher ( 95% CI 3–38% , p = 0 . 024 ) , and for each 10% increase in KT ratio , post-baseline NAb titers were 21% lower ( 95% CI 5–37% lower , p = 0 . 009 ) after adjustment for age , baseline NAb titers and HIV VL . There was no evidence for an association between CD4 or CD8 T cell count and YFV immunogenicity ( multiplicative effect per 10% increase 1 . 05 , 95% CI 0 . 91–1 . 20 , p = 0 . 469 , and 0 . 95 , 95% CI 0 . 86–1 . 04 , p = 0 . 295 , respectively ) or between Human Pegivirus co-infection and YFV immunogenicity among HIV-infected individuals ( fold change 0 . 65 , 95% CI 0 . 09–4 . 47 , p = 0 . 659 , Table 4 ) . Adjusted for age and baseline NAb titers , having detectable plasma HIV was associated with 60% lower YF-specific NAb titers ( fold change 0 . 40 , 95% CI 0 . 21–0 . 75 , p = 0 . 004 ) . As to confirm the effects of CD4/CD8 ratio and KT ratio on NAb titers , we performed simple non-parametric correlation tests; as expected , CD4/CD8 ratio correlated positively with NAb titers in all time-points , with statistically significant correlation in visit 28 ( Rho = 0 . 74 , p = 0 . 0139 ) and visit 365 ( Rho = 0 . 9 , p = 0 . 0374 ) . Similarly , KT ratio correlated negatively with NAb titers in all time-points , with statistically significant correlation in visit 56 ( Rho = -0 . 76 , p = 0 . 0171 ) .
In this prospective cohort of individuals receiving YFV , those with HIV had similar levels of YFV viremia and AEs as HIV-uninfected controls . Compared to controls , HIV-infected participants also had similar initial immunogenicity to YFV , measured by YF-specific NAb titers at 7 , 14 , 28 , 56 , and 84 days after vaccination , adjusted for age , sex and previous YFV . However , HIV status was independently associated with lower persistence of YF-specific NAb titers one year after vaccination . In the analysis of predictors of immunogenicity among HIV-infected participants , lower CD4/CD8 ratio , higher KT ratio and detectable HIV VL were associated with lower YF-specific NAb titers . There was no evidence for an association between viremia by Human Pegivirus , CD4 and CD8 T cell counts and YF-specific NAb titers among HIV-infected individuals . Earlier studies of YFV immunogenicity including HIV-infected patients in the pre-ART period or in the initial phases of ART had demonstrated that CD4 T cell count and HIV VL predicted YF-specific NAb titers [7–12] . However , in the current era of early ART initiation , more HIV-infected patients are expected to have undetectable HIV VL and higher CD4 T cell counts . In our study , despite the elevated CD4 T cell count and high proportion of ART-suppressed individuals , HIV status was still associated with lower persistence of YF-specific NAb titers following an apparently adequate initial immunogenicity . Our results are consistent with prior publications suggesting that HIV-infected subjects still present lower responses to vaccines [1 , 6 , 16] . In addition , while the START and TEMPRANO trials demonstrated that earlier ART initiation dramatically reduces the risk of infectious outcomes , there was still a substantial risk of infectious outcomes in the immediate ART arms [36 , 37] . Thus there are likely to be subtle immune defects that persist despite early ART initiation . Our study provides potentially important insights into mechanisms that might contribute to this persistent risk of infectious complications as well as point of care diagnostics that might identify patients at highest risk . For example , higher plasma KT ratio–a marker of IDO activity–strongly predicted lower YFV Nab titers after vaccination . While our observational study cannot assess causality , the fact that IDO-generated tryptophan catabolites suppress lymphocyte proliferation and function provides a plausible mechanistic pathway of its detrimental effect on vaccine responsiveness and , more broadly , adaptive immunity . Indeed , higher IDO activity has already been shown to predict increased mortality in several cohorts of ART-suppressed HIV-infected individuals [38–40] . While higher IDO activity might simply be a surrogate of other immunologic defect causally associated with impaired B cell function ( e . g . , the extent of T follicular helper cell infection and/or dysfunction in lymphoid tissues ) , a potential causal role for IDO activity in impairing vaccine responsiveness is plausible . Interestingly , high KT ratio was one of the strongest immunologic correlates of low CD4/CD8 ratio in another recent study of ART-suppressed individuals [21] , suggesting that this biomarker–already obtained as part of routine clinical care—might help identify individuals with highest risk of impaired vaccine responses and adaptive immune defects . Collectively , our findings further encourage the development of therapeutic interventions to reduce immune activation in ART-treated HIV-infected individuals [18 , 20] . Early ART initiation is a well-recognized , yet insufficient strategy to normalize persistent immune activation [41] , and additional strategies including inhibition of IDO activity are currently under study [42] . While effective interventions to inhibit immune activation are not available for clinical use , another important implication of our findings is the potential to substantiate recommendations for a booster dose of YFV for HIV-infected individuals at permanent or recurring risk of wild YF . In a recent publication , the Advisory Committee on Immunization Practices from Centers for Disease Control and Prevention published recommendations for YFV , suggesting HIV-infected individuals may benefit from a booster vaccination , which would not be recommended in routine circumstances due to the high immunogenicity and durability of YFV in the general population [43] . Our results suggest that a booster YFV may be beneficial even for HIV-infected individuals with high CD4 T cell counts . Since lower persistence of NAb was observed one year after YF vaccination , and AE following a booster dose of YFV are rare [28 , 29] , either periodic monitoring of YF-NAb or administration of a booster YFV dose could be recommended for HIV-infected individuals at permanent or recurring risk of wild YF as early as one year after primary vaccination . Additional studies are necessary to determine the durability of immunogenicity after a booster vaccination in this population . Because all included participants received YFV as indicated due to potential risk of exposure to wild-type virus , we cannot rule out that natural exposure , rather than YFV alone , partially accounted for the observed NAb titers . However , most participants were residents in non-endemic areas and received YFV due to temporary visits to endemic regions with low risk of natural exposure . Furthermore , this potential competing immune stimulus would likely occur non-differentially regarding HIV status . Consequently , we do not believe our results are substantially compromised by natural exposure to wild YF virus . Due to the small sample size , our results must be interpreted with caution . The model addressing predictors of YFV immunogenicity among HIV-infected participants included only 12 individuals followed longitudinally with 7 repeated outcome measurements . Although statistical methods for longitudinal analysis typically reduce error variance and improve power , this exploratory analysis needs confirmation in larger samples and different settings . In addition , our study was likely underpowered to provide definitive conclusions regarding the effect of HIV status on initial YFV response , on risk of AE ( in particular rare severe AE ) and on viremia by the YFV virus . Sensitivity for detection of AE was enhanced in our study by the measurement of solicited clinical AE , laboratory assessment of potential hematological or biochemical abnormalities , and systematic measurement of YFV viremia . Therefore , although not definitive , our findings provide important information on YFV clinical and laboratory adverse events , as well as vaccine virus kinetics among HIV-infected participants . Our study may also be underpowered to detect significant effects of viremia by Human Pegivirus , CD4 and CD8 T cell counts on YF-specific NAb titers among HIV-infected participants . Our study included HIV-infected individuals with a very high range of CD4 T cell count , and we cannot rule out that CD4 T cell count would still predict YFV immunogenicity among patients with wider CD4 T cell count variability . Finally , we used a single measurement of RT-PCR to determine Human Pegivirus co-infection , and were unable to discriminate recent unresolved infections from chronic infections . In conclusion , HIV-infected individuals have impaired NAb response to YFV due to a poorer persistence of antibodies , despite a seemingly normal initial response . Immune activation seems to reduce YFV immunogenicity , consistent with the observation that immune activation markers are useful predictors of clinical outcomes in the current era of HIV care [21 , 22 , 38] . A booster dose of YFV , although not recommended in routine circumstances , may be beneficial for HIV-infected individuals at permanent or recurring risk of wild YF . | Yellow Fever ( YF ) vaccine is considered one of the most effective vaccines ever produced . However , previous studies suggest that HIV impairs YF vaccine response . In this study , we assessed if HIV infection impacts the risk of adverse events and could reduce antibody response to YF vaccine . We explored if laboratory markers of persistent inflammation , frequently present among HIV-infected patients , could predict antibody response to YF vaccine in this population . We found that HIV had no significant effect on adverse events or levels of antibodies through 3 months after vaccination , but this may be limited by the small sample size of 12 HIV-infected and 45-uninfected participants in the study . However , we were able to show that , compared to HIV-uninfected participants , HIV–infected patients had lower antibody titers 1 year following YF vaccine even after statistical adjustment for the potential effects of sex , age and prior vaccination . Persistent inflammation seems to reduce YF vaccine antibody response in HIV-infected participants . In conclusion , HIV-infected individuals have impaired antibody response to YFV due to a poorer persistence of antibodies , despite a seemingly normal initial response . HIV-infected patients at permanent or recurring risk of YF infection may benefit from a booster dose of YF vaccine . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"preventive",
"medicine",
"vaccines",
"rna",
"viruses",
"vaccination",
"and",
"immunization",
"antibodies",
"antibody",
"response",
"public",
"and",
"occupational",
"health",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"booster",
"doses",
"proteins",
"medical",
"microbiology",
"hiv",
"antigens",
"microbial",
"pathogens",
"immune",
"response",
"biochemistry",
"viremia",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"lentivirus",
"organisms"
] | 2016 | CD4/CD8 Ratio and KT Ratio Predict Yellow Fever Vaccine Immunogenicity in HIV-Infected Patients |
The intracellular bacterial pathogen Legionella pneumophila causes an inflammatory pneumonia called Legionnaires' Disease . For virulence , L . pneumophila requires a Dot/Icm type IV secretion system that translocates bacterial effectors to the host cytosol . L . pneumophila lacking the Dot/Icm system is recognized by Toll-like receptors ( TLRs ) , leading to a canonical NF-κB-dependent transcriptional response . In addition , L . pneumophila expressing a functional Dot/Icm system potently induces unique transcriptional targets , including proinflammatory genes such as Il23a and Csf2 . Here we demonstrate that this Dot/Icm-dependent response , which we term the effector-triggered response ( ETR ) , requires five translocated bacterial effectors that inhibit host protein synthesis . Upon infection of macrophages with virulent L . pneumophila , these five effectors caused a global decrease in host translation , thereby preventing synthesis of IκB , an inhibitor of the NF-κB transcription factor . Thus , macrophages infected with wildtype L . pneumophila exhibited prolonged activation of NF-κB , which was associated with transcription of ETR target genes such as Il23a and Csf2 . L . pneumophila mutants lacking the five effectors still activated TLRs and NF-κB , but because the mutants permitted normal IκB synthesis , NF-κB activation was more transient and was not sufficient to fully induce the ETR . L . pneumophila mutants expressing enzymatically inactive effectors were also unable to fully induce the ETR , whereas multiple compounds or bacterial toxins that inhibit host protein synthesis via distinct mechanisms recapitulated the ETR when administered with TLR ligands . Previous studies have demonstrated that the host response to bacterial infection is induced primarily by specific microbial molecules that activate TLRs or cytosolic pattern recognition receptors . Our results add to this model by providing a striking illustration of how the host immune response to a virulent pathogen can also be shaped by pathogen-encoded activities , such as inhibition of host protein synthesis .
In metazoans , the innate immune system senses infection through the use of germline-encoded pattern recognition receptors ( PRRs ) that detect pathogen-associated molecular patterns ( PAMPs ) , such as lipopolysaccharide or flagellin [1] . PAMPs are conserved molecules that are found on non-pathogenic and pathogenic microbes alike , and consequently , even commensal microbes are capable of activating PRRs [2] . Thus , it has been proposed that additional innate immune mechanisms may exist to discriminate between pathogens and non-pathogens [3] , [4] . In plants , selective recognition of pathogens is accomplished by detection of the enzymatic activities of “effector” molecules that are delivered specifically by pathogens into host cells . Typically , the effector is an enzyme that disrupts host cell signaling pathways to the benefit of the pathogen . Host sensors monitoring or “guarding” the integrity of the signaling pathway are able to detect the pathogen-induced disruption and initiate a protective response . This mode of innate recognition is termed “effector-triggered immunity” [5] and represents a significant component of the plant innate immune response . It has been suggested that innate recognition of pathogen-encoded activities , which have been termed “patterns of pathogenesis” in metazoans [3] , could act in concert with PRRs to distinguish pathogens from non-pathogens , leading to qualitatively distinct responses that are commensurate with the potential threat . However , few if any examples of “patterns of pathogenesis” have been shown to elicit innate responses in metazoans . The gram negative bacterial pathogen Legionella pneumophila provides an excellent model to address whether metazoans respond to pathogen-encoded activities in addition to PAMPs . L . pneumophila replicates in the environment within amoebae [6] , but can also replicate within alveolar macrophages in the mammalian lung [7] , where it causes a severe inflammatory pneumonia called Legionnaires' Disease [6] . Because its evolution has occurred primarily or exclusively in amoebae , L . pneumophila appears not to have evolved significant immune-evasive mechanisms . Indeed , most healthy individuals mount a robust protective inflammatory response to L . pneumophila , resulting from engagement of multiple redundant innate immune pathways [8] . We hypothesize , therefore , that as a naïve pathogen , L . pneumophila may reveal novel innate immune responses that better adapted pathogens may evade or disable [9] . In host cells , L . pneumophila multiplies within a specialized replicative vacuole , the formation of which is orchestrated by bacterial effector proteins translocated into the host cytosol via the Dot/Icm type IV secretion system [10] . In addition to its essential roles in bacterial replication and virulence , the Dot/Icm system also translocates bacterial PAMPs , such as flagellin , nucleic acids , or fragments of peptidoglycan , that activate cytosolic immunosurveillance pathways [8] , [11] , [12] , [13] , [14] , [15] , [16] . There are also recent suggestions in the literature that Dot/Icm+ L . pneumophila may stimulate additional , uncharacterized immunosurveillance pathways [8] , [17] . Overall , the molecular basis of the host response to Dot/Icm+ L . pneumophila remains poorly understood . Here we show that macrophages infected with virulent L . pneumophila make a unique transcriptional response to a bacterial activity that disrupts a vital host process . We show that this robust transcriptional response requires the Dot/Icm system , and cannot be explained solely by known PAMP-sensing pathways . Instead , we provide evidence that the response requires the enzymatic activity of five secreted bacterial effectors that inhibit host protein synthesis . Effector-dependent inhibition of protein synthesis synergized with PRR signaling to elicit the full transcriptional response to L . pneumophila . The response to the bacterial effectors could be recapitulated through the use of pharmacological agents or toxins that inhibit host translation , administered in conjunction with a PRR agonist . Thus , our results provide a striking example of a host response that is shaped not only by PAMPs but also by a complementary “effector-triggered” mechanism that represents a novel mode of immune responsiveness in metazoans .
We initially sought to identify host responses that discriminate between pathogenic and non-pathogenic bacteria . Our strategy was to compare the host response to wildtype virulent L . pneumophila with the host response to an avirulent L . pneumophila mutant , ΔdotA . ΔdotA mutants lack a functional Dot/Icm secretion system , and thus fail to translocate effectors into the host cytosol , but they nevertheless express the normal complement of PAMPs that engage Toll-like receptor pathways . We performed transcriptional profiling experiments on macrophages infected with either wildtype L . pneumophila or the avirulent ΔdotA mutant . In the microarray experiments , Caspase-1−/− macrophages were used to eliminate flagellin-dependent macrophage death , which would otherwise differ between wildtype and ΔdotA infections [12] , [14] , [16] , but our results were later validated with wildtype macrophages ( see below ) . RNA was collected from macrophages at a timepoint when there were similar numbers of bacteria in both wildtype-infected and ΔdotA-infected macrophages . Microarray analysis revealed 166 genes that were differentially induced >2-fold in a manner dependent on type IV secretion ( Figure 1A and Table S1 ) . The induction of some of the Dot/Icm-dependent genes , e . g . Ifnb , could be explained by cytosolic sensing pathways that have been previously characterized [11] , [13] , [18] . However , much of the response to Dot/Icm+ bacteria did not appear to be accounted for by host pathways known to recognize L . pneumophila . For reasons discussed below , we refer specifically to this unexplained Dot/Icm-dependent transcriptional signature as the ‘effector-triggered response , ’ or ETR . The ETR includes many genes thought to be important for innate immune responses , including the cytokines/chemokines Csf1 , Csf2 , Ccl20 , and Il23a; the surface markers Sele , Cd83 , and Cd44; and the stress response genes Gadd45 , Egr1 , and Egr3 . Other ETR targets were genes whose function in macrophages has not been determined ( e . g . , Gem , which encodes a small GTPase ) ( Figure 1A and Table S1 ) . We selected several of the most highly induced genes for validation by quantitative reverse-transcription PCR . We confirmed that Il23a , Csf2 and Gem transcripts were induced 100 to >1000-fold more by pathogenic wildtype L . pneumophila as compared to the ΔdotA mutant ( Figure 1B ) . In subsequent experiments we focused on these three genes , as they provided a sensitive readout of the ETR . To assess whether the ETR might be important during L . pneumophila infection in vivo , we infected B6 and Il23a−/− mice intranasally with L . pneumophila . Il23a−/− mice displayed a significant defect in host cell recruitment to the lungs 24 hours after infection ( Figure 1C ) , consistent with the known role of IL-23 in neutrophil recruitment to sites of infection [19] . The phenotype of Il23a−/− mice was not due to decreased bacterial burden in these mice ( Figure 1C ) . Thus at least one transcriptional target of the ETR plays a role in the host response , though there are clearly numerous redundant pathways that recognize L . pneumophila in vivo [8] . In order to identify the host pathway ( s ) responsible for induction of the ETR , we first examined innate immune pathways known to recognize L . pneumophila . Induction of the representative genes Il23a , Csf2 , and Gem did not require the previously described Naip5/Nlrc4 flagellin-sensing pathway [20] , as infection with a flagellin-deficient mutant ( ΔflaA ) also induced robust expression of these genes ( Figure 1A , B and Table S2 ) . Moreover , Il23a , Csf2 and Gem were strongly ( >1000-fold ) induced in the absence of the Mavs/Irf3/Irf7 signaling axis shown previously to respond to L . pneumophila [11] , [13] , [18] ( Figure 1D , and data not shown ) . As suggested by previous transcriptional profiling experiments [17] , we confirmed that Myd88−/−and Rip2−/−macrophages , which are defective in TLR and Nod1/Nod2 signaling , respectively , strongly upregulated Il23a and Gem following infection with wildtype L . pneumophila ( Figure 2A ) . Induction of Il23a was abrogated in Myd88−/−Rip2−/− and Myd88−/−Nod1−/−Nod2−/− macrophages; however , these macrophages still robustly induced Gem ( Figure 2A , and data not shown ) . These data indicate that TLR/Nod signaling is necessary for induction of some , but not all , genes in the ETR . Furthermore , the intact induction of Gem in Myd88−/−Nod1−/−Nod2−/− macrophages implies the existence of an additional pathway . To address the further question of whether TLR/Nod signaling was sufficient for induction of the ETR , we treated uninfected macrophages with synthetic TLR2 and/or Nod2 ligands ( Pam3CSK4 and MDP , respectively ) . These ligands did induce low levels of Il23a , but could not recapitulate the robust ( 100–1000 fold ) upregulation indicative of the ETR ( Figure 2B ) . The defective induction of ETR target genes was not due to inefficient delivery of the ligands , as Pam3CSK4 and MDP were able to strongly induce Il1b ( Figure 2B ) . To confirm this result in a more physiologically relevant system , we infected macrophages with the Gram-positive intracellular bacterial pathogen Listeria monocytogenes , which is known to activate both TLRs and Nods [21] . Infection with L . monocytogenes resulted only in weak Il23a induction ( ∼50 fold less than wildtype L . pneumophila at the same initial multiplicity of infection ) ( Figure 2C ) . A failure to strongly upregulate Il23a did not appear to be due to poor infectivity of L . monocytogenes , since the cytosolically-induced gene Ifnb [21] was robustly transcribed ( Figure 2C ) . Taken together , these results suggest that TLR/Nod signaling , while necessary for transcription of some ETR targets , is not sufficient to account for the full induction of the ETR by L . pneumophila . Though PRRs do play some role in induction of the ETR , we could not identify a known PAMP-sensing pathway that fully accounted for this robust transcriptional response . Therefore we considered the hypothesis that host cells respond to an L . pneumophila-encoded activity in addition to PAMPs . Since L . pneumophila manipulates host cell biology via its Dot/Icm-secreted effectors , we analyzed the transcriptional response of macrophages infected with L . pneumophila ΔicmS/ΔicmW mutants , which express a functional Dot/Icm system [15] , but lack chaperones required for secretion of many effectors . Macrophages infected with ΔicmS/ΔicmW L . pneumophila exhibited a ∼50-fold defect in induction of Il23a and Gem ( Figure 2D ) . Thus , secreted effectors ( or the physiological stresses they impart ) appear to participate in induction of the ETR . To identify potential host pathways capable of inducing ETR target genes , we treated macrophages with known inducers of host cell stress responses . We found that the pharmacological agents thapsigargin and tunicamycin , which inhibit host translation via induction of endoplasmic-reticulum ( ER ) stress [22] , synergized with a TLR2 ligand to induce high levels of Il23a and Gem ( Figure 2E , and data not shown ) . To test whether L . pneumophila might elicit the ETR via induction of ER stress , we measured Xbp-1 splicing and transcription of classical ER stress markers in macrophages infected with L . pneumophila . However , we found no evidence of ER stress in these macrophages ( data not shown ) . Instead , we considered the possibility that thapsigargin induces the ETR through inhibition of protein synthesis . In fact , the L . pneumophila Dot/Icm system was previously reported to translocate several effector enzymes that inhibit host translation [23] , [24] , [25] . Therefore we hypothesized that inhibition of host protein synthesis by L . pneumophila [26] might be responsible for induction of the ETR . To determine whether inhibition of host translation by L . pneumophila was critical for induction of the ETR , we generated a mutant strain of L . pneumophila , called Δ5 , which lacks five genes encoding effectors that inhibit host translation ( lgt1 , lgt2 , lgt3 , sidI , sidL; Figure S1; Table S3 ) . Three of these effectors ( lgt1 , lgt2 , lgt3 ) , which share considerable sequence homology , are glucosyltransferases that modify the mammalian elongation factor eEF1A and block host translation both in vitro and in mammalian cells [23] , [25] . A fourth effector ( sidI ) binds both eEF1A and another host elongation factor , eEF1Bγ , and has also been shown to inhibit translation in vitro and in cells infected with L . pneumophila [24] . The fifth effector , sidL , is toxic to mammalian cells and is capable of inhibiting protein translation in vitro via an unknown mechanism ( data not shown ) . Moreover , its expression by L . pneumophila enhances global translation inhibition in infected macrophages ( see below ) . These 5 effectors appear to be important for survival within the pathogen's natural host , since the Δ5 mutant displayed a ∼10-fold growth defect in Dictyostelium amoebae ( Figure 3A ) . By contrast , the Δ5 mutant showed no growth defect in macrophages ( Figure 3B ) , but was defective , compared to wildtype , in its ability to inhibit host protein synthesis ( Figure 3C ) . Although to a lesser degree than wildtype bacteria , the Δ5 mutant still appears to partially inhibit host protein synthesis , suggesting that L . pneumophila may encode additional inhibitors of host translation . Nevertheless , macrophages infected with Δ5 exhibited striking defects in induction of the ETR , including a ∼50-fold defect in induction of Il23a , Gem , and Csf2 ( Figure 3D and Table S4 ) . Importantly , the Dot/Icm-dependent induction of Ifnb , which is induced via a separate pathway [11] , [13] , [15] , remained intact ( Figure 3D ) , implying that the Δ5 mutant was competent for infection and Dot/Icm function . Individual deletion mutants of each of the five effectors showed no defect in Il23a , Csf2 , or Gem induction , whereas a mutant lacking four of the five ( Δlgt1Δlgt2Δlgt3ΔsidI ) had a partial defect ( Figure 3D , and data not shown ) . Complementation of Δ5 with wildtype lgt2 or lgt3 restored induction of Il23a and Gem , but complementation with mutant lgt2 or lgt3 lacking catalytic activity did not ( Figure 3E ) . These results are significant because they show that macrophages make an innate response to a pathogen-encoded activity and that recognition of the effector molecules themselves is not likely to explain the ETR . We then tested more directly whether the ETR was induced by translation inhibition . The defective induction of Il23a , Csf2 , and Gem in macrophages infected with ΔdotA or Δ5 was rescued by addition of the translation inhibitor cycloheximide ( Figure 4A , and data not shown ) . These results support the hypothesis that induction of the ETR by L . pneumophila involves inhibition of translation by the five deleted effectors . Importantly , the potent induction of Il23a , Csf2 and Gem by L . pneumophila could be recapitulated in uninfected macrophages by treatment with the translation elongation inhibitors cycloheximide ( Figure 4B ) or puromycin ( Figure 4C ) , or the initiation inhibitor bruceantin ( Figure 4D ) , in conjunction with the TLR2 ligand Pam3CSK4 . These three translation inhibitors possess different targets and modes of action , making it unlikely that the common host response to each of them is due to nonspecific drug effects . Thus , translation inhibition in the context of TLR signaling provokes a specific transcriptional response . Translation inhibitors alone were capable of inducing some , but not all , effector-triggered transcriptional targets ( Figure 4B , C , and D ) , supporting our model that translation inhibition acts in concert with classical PRR signaling to generate the full effector-dependent signature . Microarray analysis indicated that the five effectors accounted for induction of at least 54 ( ∼30% ) of the Dot/Icm-dependent genes ( Figure 5A and Table S4 ) . We investigated how inhibition of protein synthesis by L . pneumophila might elicit a host response . Although translation inhibition by cycloheximide has long been reported to induce cytokine production [27] , the mechanism by which it acts remains poorly understood . Since the induction of Il23a and Csf2 is NF-κB dependent ( [28] , and data not shown ) , we examined a role for this pro-inflammatory transcription factor in induction of these ETR targets . NF-κB is normally suppressed by its labile inhibitor IκB , which is ubiquitinated and degraded in response to TLR and other inflammatory stimuli . IκB is itself a target of NF-κB-dependent transcription , and resynthesis of IκB is critical for the homeostatic termination of NF-κB signaling . In the absence of protein synthesis , we hypothesized that IκB may fail to be resynthesized as it turns over , thereby permitting continued NF-κB activity . To test this hypothesis , we measured IκB levels in infected macrophages over time . We observed a prolonged decrease in levels of IκB protein in macrophages infected with wildtype L . pneumophila , consistent with previous observations [17] ( Figure 5B ) . In contrast , infection with Δ5 triggered only a transient loss of IκB , similar to infection with the secretion-deficient ΔdotA mutant ( Figure 5B ) . The Δ5 mutant could induce sustained IκB degradation when complemented with plasmid-encoded lgt3 , but not with a mutant effector lacking glucosyltransferase activity ( Figure 5C ) , demonstrating that the sustained loss of IκB is due to the activity of the bacterial effector . To confirm that the prolonged loss of IκB did indeed result in sustained NF-κB activation , we measured NF-κB translocation to the nucleus in macrophages infected with wildtype , ΔdotA , or Δ5 L . pneumophila . While all three strains initially induced nuclear translocation of NF-κB , at later timepoints we observed decreased levels of nuclear NF-κB in macrophages infected with the ΔdotA or Δ5 strains compared to those infected with wildtype L . pneumophila ( Figure 5D ) . Thus , translation inhibition by the 5 effectors results in sustained loss of IκB and enhanced activation of NF-κB . NF-κB signaling is also inhibited by other de novo expressed proteins such as A20 [29] . We therefore used A20−/− macrophages , which exhibit prolonged NF-κB activation in response to TLR signaling [29] , to further test the hypothesis that sustained NF-κB signaling can induce targets of the ETR . Strikingly , we found that the defective induction of Il23a and Csf2 by Δ5 was rescued in A20−/− macrophages ( Figure 5E ) . Taken together , these observations suggest a model in which disrupted protein synthesis , and the subsequent failure to synthesize inhibitors of NF-κB signaling ( e . g . IκB and A20 ) , leads to sustained activation of NF-κB ( Figure 6 ) . In turn , we suggest that this prolonged activation of NF-κB results in enhanced transcription of a specific subset of genes . Importantly , sustained NF-κB activation did not appear to result in transcriptional superinduction of all NF-κB-dependent target genes . Microarray analysis ( Figure 5A and Table S4 ) suggested that only a subset of NF-κB-induced genes was preferentially induced by translation inhibition . For example , Nfkbia ( encoding IκBα ) , a known NF-κB target gene , was not dramatically superinduced by wildtype compared to Δ5 L . pneumophila ( Figure 5F ) . The molecular mechanism that results in specific superinduction of certain NF-κB-dependent target genes is not yet clear and may be complex ( see Discussion ) . Inhibition of protein synthesis by L . pneumophila may also result in activation of other synergistic signaling pathways [30] , such as MAP kinases ( [17] , data not shown ) , or in mRNA stabilization . In light of these possibilities , we confirmed that the increase in expression of ETR target genes does involve de novo transcription , by quantifying transcript levels using primers specific for unspliced mRNA ( Figure S2A ) . We also tested whether mRNA stabilization contributed to induction of the ETR by infecting macrophages in the presence of the transcription inhibitor actinomycin D and quantifying ETR target mRNAs at successive timepoints . Our results suggested that RNA stabilization does not play a major role in induction of these particular ETR targets ( Figure S2B ) , though we do not rule it out as a possible mechanism for increasing some mRNA transcripts in the ETR . Although inhibition of protein synthesis potently induces transcription of certain target genes , a central question is whether this transcriptional response is sufficient to overcome the translational block , and result in increased protein production . Accordingly , we measured the protein levels of GM-CSF ( encoded by the Csf2 gene ) in the supernatant of infected macrophages . GM-CSF protein was preferentially produced by cells infected with wildtype L . pneumophila as compared to cells infected with Δ5 ( Figure 7A ) . The defect in cytokine production by Δ5-infected macrophages was not due to poor bacterial growth ( Figure 3B ) , increased cytotoxicity ( Figure S3A ) , or defective secretion ( Figure S3B ) , and could be rescued by addition of cycloheximide ( Figure 7A ) . Thus translation inhibition can paradoxically lead to increased production of certain proteins , perhaps because transcriptional superinduction of specific transcripts is sufficient to overcome the partial translational block mediated by L . pneumophila . We did not observe defects in cytokine induction or altered bacterial replication in B6 mice infected with the Δ5 mutant . This is perhaps not surprising , since many redundant innate immune signaling pathways are known to recognize and restrict the growth of L . pneumophila in vivo [8] . Indeed , we found that dendritic cells infected with L . pneumophila upregulate ETR target genes independently of the Dot/Icm secretion system ( Figure S4 ) , and hence translation inhibition appears not to be essential for their response to L . pneumophila . However , many other pathogens also produce toxins that inhibit host protein synthesis ( e . g . , Diphtheria Toxin , Shiga Toxin , Pseudomonas Exotoxin A ) . Thus , to test whether translation inhibition may be a general stimulus that acts with PRRs to elicit a host response to diverse pathogens , we treated uninfected macrophages with Diphtheria Toxin ( DT ) or Exotoxin A ( ExoA ) in conjunction with a TLR2 ligand . Importantly , both of these toxins inhibit translation by ADP-ribosylation of eEF2 , a mechanism of action distinct from that employed by the five L . pneumophila effectors . When administered with Pam3CSK4 , both toxins robustly induced Il23a ( Figure 7B ) . DT alone was sufficient to induce Il23a , most likely due to the presence of TLR ligands in the recombinant protein preparation . Consistent with these findings , Shiga Toxin , which inhibits translation by yet another mechanism , has also been reported to superinduce cytokine responses in a cultured cell line [31] . The existence of a common host response to diverse mechanisms of translation inhibition provides strong evidence that host cells can specifically respond to this disruption of their physiology , in addition to recognizing microbial molecules . Finally , since in vivo infection with L . pneumophila results in multiple redundant responses that may have obscured our ability to detect an in vivo phenotype for the Δ5 mutant , we turned to a simpler model to ascertain whether the ETR can be induced in vivo . In this model , purified Exotoxin A was administered intranasally to inhibit host protein synthesis in the lungs . Importantly , we found that translation inhibition appears to synergize with TLRs to elicit an immune response in vivo , as mice treated intranasally with ExoA and Pam3CSK4 produced significant amounts of the characteristic effector-triggered cytokine GM-CSF ( Figure 7C ) . Consistent with our observations in vitro ( Figure 7B ) , intranasal instillation of ExoA or Pam3CSK4 individually resulted in a much more modest response , providing further evidence that two signals—PRR activation and translation inhibition—are needed to generate the full effector-dependent signature . ExoA alone was sufficient to induce transcription of Gem and Csf2 mRNA in the lung ( Figure S5 ) , again in agreement with in vitro observations that translation inhibition alone can induce transcription of some target genes ( Figure 4B , C , and D ) . Taken together , our results demonstrate that translation inhibition by multiple pathogens can lead to a common innate response in cultured cells and in vivo .
In this study , we have demonstrated that inhibition of host translation by bacterial effectors or toxins can elicit a potent response from the host . We thus provide strong evidence for a model of innate immune recognition that is complementary to , but distinct from , the classic PAMP-based model . Most notably , we show that the immune system can mount a response to a pathogen-associated activity , in addition to pathogen-derived molecules . In our model , it is important to emphasize that there is no need for a specific host receptor or sensor per se . Instead , our data support the hypothesis that a pathogen-mediated block in the synthesis of short-lived host signaling inhibitors ( e . g . IκB , A20 ) results in the sustained activation of an inflammatory mediator ( e . g . NF-κB ) ( Figure 6 ) . As such , our model more closely resembles the indirect “guard” type mechanisms that plants utilize , in conjunction with PRRs , to sense pathogens [5] . The labile nature of IκB makes it an effective “guard” to monitor the integrity of host translation , since the short half-life of this protein ensures that its abundance will decrease quickly during conditions where translation is inhibited . There are growing suggestions that host responses to ‘patterns of pathogenesis’ [3] , or harmful pathogen-associated activities , may indeed comprise a general innate immunosurveillance strategy in metazoans . For example , ion channel formation by influenza virus appears to activate the Nlrp3 inflammasome [32] , and Salmonella effectors that stimulate Rho-family GTPases appear to trigger specific inflammatory responses [33] . However , in these examples , both the precise host cell disruption and the mechanism by which the host responds remain unclear . Our results are significant because we have provided a mechanism by which host cells generate a unique transcriptional response to a specific pathogen-encoded activity , namely , inhibition of host protein synthesis . An important question is whether the innate response to translation inhibition represents a host strategy for detecting and containing a pathogen , or is rather a manipulation of the host immune system by the bacterium . Given the natural history of L . pneumophila , we consider it unlikely that this pathogen has evolved to manipulate the innate immune system [9] . L . pneumophila is not thought to be transmitted among mammals; instead , our data ( Figure 3A ) suggest that the five effectors described here probably evolved to aid survival in amoebae , the natural hosts of L . pneumophila . We therefore favor the hypothesis that the innate immune system has evolved to respond to disruptions in protein translation , an essential activity that is targeted by multiple viral and bacterial pathogens . We observed that inhibition of translation in the context of PRR signaling results in the transcriptional superinduction of a specific subset of >50 genes , including Il23a , Gem , and Csf2 , that constitute an ‘effector-triggered’ response . We propose that at least some of these genes are superinduced upon the sustained activation of transcription factors such as NF-κB , although it is important to emphasize that the host response to protein synthesis inhibition is complex and likely involves other pathways as well , such as MAP kinase activation ( data not shown ) . Interestingly , we observed that not all NF-κB-dependent target genes are superinduced by translation inhibition . For example , Nfkbia ( encoding the IκB protein ) was not superinduced in wildtype L . pneumophila infection ( Figure 5F ) . This selective superinduction of certain target genes may be significant , since it allows the host to respond to a pathogen-dependent stress by altering not only the magnitude but also the composition of the transcriptional response . Moreover , if IκB were superinduced , this would presumably act to reverse or prevent sustained NF-κB signaling , resulting in little net gain . The mechanism by which prolonged NF-κB signaling may preferentially enhance transcription of the specific subset of effector-triggered genes is not yet clear . However , recent studies have shown that the chromatin context for several of these genes ( e . g . , Il23a , Csf2 ) is in a relatively ‘closed’ conformation [34] , [35] . This may render the genes refractory to strong transcriptional induction under a normal TLR stimulus , but enable them to become highly induced upon prolonged NF-κB activation . It is interesting to note that genes such as Il23a and Csf2 are classified as ‘primary’ response genes [34] , [35] simply because they are inducible in the presence of cycloheximide . What is not often discussed is the possibility , demonstrated here , that inhibition of protein synthesis by cycloheximide is a key stimulus that induces transcription of these genes . The consequences of the host response to translation inhibition are likely to be difficult to measure in the context of a microbial infection in vivo . Presumably , most pathogens that disrupt host translation derive benefit from this activity , perhaps by increasing availability of amino acid nutrients or by dampening production of the host response . These benefits may be offset by an enhanced host response to translation inhibition itself . It is possible that the robust innate immune response to translation inhibition serves primarily to compensate for the decrease in translation , resulting in little net change in the output of the immune response . Accordingly , the lack of an apparent phenotype during in vivo infection with Δ5 may reflect the sum of multiple positive and negative effects that result from translation inhibition . Additionally , as suggested by our data ( Figure S4 ) the response to L . pneumophila in vivo may involve non-macrophage cell types in which translation inhibition does not play a crucial role . While PRR-based sensing of microbial molecules is certainly a fundamental mode of innate immune recognition , it is not clear how PRRs alone might be able to distinguish pathogens from non-pathogens , and thereby mount responses commensurate with the potential threat . Our results demonstrate that pathogen-mediated interference with a key host process ( i . e . , host protein synthesis ) , in concert with PRR signaling , results in an immune response that is qualitatively distinct from the response to an avirulent microbe . Although induction of some genes in the ETR ( e . g . , Gem ) occurs in response to inhibition of protein synthesis alone , much of the ETR is due to the combined effects of PAMP recognition and effector-dependent inhibition of protein synthesis . A requirement for two signals might be rationalized by the fact that the ETR includes potent inflammatory cytokines such as GM-CSF or IL-23 , which can drive pathological inflammation [36] and autoimmunity [37] if expressed inappropriately . Restricting production of potentially dangerous cytokines to instances where a pathogenic microbe is present may be a strategy by which hosts avoid self-damage unless necessary for self-defense . Thus , we propose that the host response to a harmful pathogen-encoded activity may represent a general mechanism by which the immune systems of metazoans distinguish pathogens from non-pathogens .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Animal Care and Use Committee at the University of California , Berkeley ( Protocol number R301-0311BCR ) . Macrophages were derived from the bone marrow of the following mouse strains: C57BL/6J ( Jackson Labs ) , A20−/− ( A . Ma , UCSF ) , Caspase-1−/− ( M . Starnbach , Harvard Medical School ) , Mavs−/− ( Z . Chen , University of Texas SW ) , Irf3/Irf7−/− ( K . Fitzgerald , U . Mass Medical School ) , Myd88−/− ( G . Barton , UC Berkeley ) , Rip2−/− ( M . Kelliher , U . Mass Medical School ) , Myd88−/−Rip2−/− ( C . Roy , Yale University ) , and Myd88−/−Nod1−/−Nod2−/− ( generated from crosses at UC Berkeley ) . Il23a−/− mice were from N . Ghilardi ( Genentech ) . Macrophages were derived from bone marrow by 8d culture in RPMI supplemented with 10% serum , 100 µM streptomycin , 100 U/mL penicillin , 2 mM L-glutamine , and 10% supernatant from 3T3-M-CSF cells , with feeding on day 5 . Dendritic cells were derived from B6 bone marrow by 6d culture in RPMI supplemented with 10% serum , 100 µM streptomycin , 100 U/mL penicillin , 2 mM glutamine , and recombinant GM-CSF ( 1:1000 , PeproTech ) . Dictyostelium discoideum amoebae were cultured at 21°C in HL-5 medium ( 0 . 056 M glucose , 0 . 5% yeast extract , 0 . 5% proteose peptone , 0 . 5% thiotone , 2 . 5 mM Na2HPO4 , 2 . 5 mM KH2PO4 , pH 6 . 9 ) . The L . pneumophila wildtype strain LP02 is a streptomycin-resistant thymidine auxotroph derived from L . pneumophila LP01 . The ΔdotA , ΔflaA , ΔicmS and ΔicmW mutants have been described [14] , [15] . Mutants lacking one or more effectors were generated from LP02 by sequential in-frame deletion using the suicide plasmid pSR47S as described [24] . Sequences of primers used for constructing deletion plasmids are listed in Table S3 . Mutants were complemented with the indicated effectors expressed from the L . pneumophila sidF promoter in the plasmid pJB908 , which encodes thymidine synthetase as a selectable marker . L . monocytogenes strain 10403S and the isogenic Δhly mutant have been described [21] . Macrophage RNA from 1 . 5×106 cells ( 6 well dishes ) was isolated using the Ambion RNAqueous Kit ( Applied Biosystems ) and amplified with the Ambion Amino Allyl MessageAmp II aRNA Amplification Kit ( Applied Biosystems ) according to the manufacturer's protocol . Microarrays were performed as described [38] . Briefly , spotted microarrays utilizing the MEEBO 70-mer oligonucleotide set ( Illumina ) were printed at the UCSF Center for Advanced Technology . Microarray probes were generated by coupling amplified RNA to Cy dyes . After hybridization , arrays were washed , scanned on a GenePix 4000B Scanner ( Molecular Devices ) , and gridded using SpotReader software ( Niles Scientific ) . Analysis was performed using the GenePix Pro 6 and Acuity 4 software packages ( Molecular Devices ) . Two independent experiments were performed . Microarray data have been deposited in the Gene Expression Omnibus database ( http://www . ncbi . nlm . nih . gov/geo/ ) under the accession number GSE26491 . Macrophages were plated in 6 well dishes at a density of 1 . 5×106 cells per well and infected at an MOI of 1 by centrifugation for 10 min at 400× g , or were treated with puromycin , thapsigargin , tunicamycin , cycloheximide ( all Sigma ) , Exotoxin A ( List Biological Labs ) , transfected synthetic muramyl-dipeptide ( MDP ) ( CalBiochem ) , or a synthetic bacterial lipopeptide ( Pam3CSK4 ) ( Invivogen ) . Dendritic cells were plated at a density of 106 cells per well and infected at an MOI of 2 as described above . Lipofectamine 2000 ( Invitrogen ) was used for transfections . Bruceantin was the kind gift of S . Starck and N . Shastri ( UC Berkeley ) , who obtained it from the National Cancer Institute , NIH ( Open Repository NSC165563 ) . A fusion of diphtheria toxin to the lethal factor translocation signal ( LFn-DT ) was the gift of B . Krantz ( UC Berkeley ) and was delivered to cells via the pore formed by anthrax protective antigen ( PA ) as described [39] . Macrophage RNA was harvested 4-6 hours post infection , as indicated , and isolated with the RNeasy kit ( Qiagen ) according to the manufacturer's protocol . RNA samples were treated with RQ1 DNase ( Promega ) prior to reverse transcription with Superscript III ( Invitrogen ) . cDNA reactions were primed with poly dT for measurement of mature transcripts , and with random hexamers ( Invitrogen ) for measurement of unspliced transcripts . Quantitative PCR was performed as described [13] using the Step One Plus RT PCR System ( Applied Biosystems ) with Platinum Taq DNA polymerase ( Invitrogen ) and EvaGreen ( Biotium ) . Transcript levels were normalized to Rps17 . Primer sequences are listed in Table S5 . Macrophages were infected in 6-well dishes at an MOI of 1 , as described above . The transcription inhibitor Actinomycin D ( 10 µg/mL , Sigma ) was added 4 hours post infection . RNA was harvested at successive timepoints and levels of indicated transcripts were assessed by quantitative RT-PCR . Age- and sex-matched B6 or Il23a−/− mice were anesthetized with ketamine and infected intranasally with 2×106 LP01 in 20 µL PBS essentially as described [13] , or were treated with ExoA or Pam3CSK4 in 25 µL PBS . Bronchoalveolar lavage was performed 24 hours post infection by introducing 800 µL PBS into the trachea with a catheter ( BD Angiocath 18 g , 1 . 3×48 mm ) . Lavage fluid was analyzed by ELISA . Total host cells in the lavage were counted on a hemocytometer . For RT-PCR experiments , all lavage samples receiving identical treatments were pooled , and RNA was isolated from the pooled cells using the RNeasy Kit as described above . FACS analysis of lavage samples labeled with anti-GR-1-PeCy7 and anti-Ly6G-PE ( eBioscience ) indicated that most cells in lavage were neutrophils . CFU were enumerated by hypotonic lysis of host cells in the lavage followed by plating on CBYE plates . Macrophages were plated in 6 well dishes at a density of 2×106 cells per well and infected at an MOI of 2 . For whole cell extract , cells were lysed in RIPA buffer supplemented with 2 mM NaVO3 , 1 mM PMSF , 1 mM DTT , and 1 X Complete Protease Inhibitor Cocktail ( Roche ) . For nuclear translocation experiments , nuclear and cytosolic fractions were obtained using the NE-PER kit ( Pierce ) according to the manufacturer's protocol . Protein levels were normalized using the micro-BCA kit ( Pierce ) and then separated on 10% NuPAGE bis-tris gels ( Invitrogen ) . Proteins were transferred to PVDF membranes and immunoblotted with antibodies to IκBα , NF-κB p65 , lamin-B or β-actin ( all Santa Cruz ) . Macrophages were plated in 24 well dishes at a density of 5×105 cells per well and infected at an MOI of 1 . After 24 h , supernatants were collected , sterile-filtered , and analyzed by ELISA using paired GM-CSF antibodies ( eBioscience ) . For quantification of intracellular GM-CSF , ELISAs were performed using cytoplasmic extract of macrophages infected for 6 h with the indicated strains . Levels of GM-CSF were normalized to total protein concentration . Recombinant GM-CSF ( eBioscience ) was used as a standard . Intracellular bacterial growth of wildtype and mutant L . pneumophila was evaluated in A/J macrophages as described [24] . D . discoideum was plated into 24-well plates at a density of 5×105 cells per well in MB medium ( modified HL-5 medium , without glucose and with 20 mM MES buffer ) three hours before infection with the indicated L . pneumophila strains at an MOI of 0 . 05 . The plates were spun at 1000 rpm for 5 minutes and incubated at 25°C . After two hours , wells were washed 3X with PBS to synchronize the infection . At successive time points , infected cells were lysed with 0 . 2% saponin and bacterial growth was determined by plating on growth medium . 2×106 macrophages were seeded in 6-well plates and infected with bacterial strains at an MOI of 2 . After 2 . 5 h , the infected cells were incubated with 1 µCi 35S-methionine ( Perkin Elmer ) in RPMI-met ( Invitrogen ) . After chase-labeling for an hour , the cells were washed 3× with PBS , lysed with 0 . 1% SDS and precipitated with TCA [24] . The protein precipitates were filtered onto 0 . 45 mm Millipore membranes and washed twice with PBS . Retained 35S was determined by a liquid scintillation counter . Macrophages were plated in 96 well dishes at a density of 5×104 cells per well and infected at an MOI of 1 . At successive timepoints , Neutral Red ( Sigma ) was added to a final concentration of 1% and incubated for 1 h . Cells were then washed with PBS , photographed , and counted [14] . | In animals , the innate immune system senses infection primarily through detection of conserved microbial molecules . It has been suggested , but not clearly established , that the immune system may also respond to pathogen-associated activities—i . e . , the manipulations of host cell processes that a pathogen employs to survive and replicate in its host . Previous studies have established that macrophages infected with the bacterial pathogen Legionella pneumophila can discriminate between virulent wildtype bacteria and an avirulent , nonreplicating mutant . Here we show that a unique host transcriptional response to virulent L . pneumophila is due to the activity of secreted bacterial proteins that inhibit host translation . Furthermore , we show that multiple bacterial toxins or chemicals that inhibit host translation can cooperate with host sensors of microbial molecules to induce the unique transcriptional response , even in the absence of bacterial infection . By demonstrating that the host mounts a response to a pathogen-encoded activity , we provide evidence for a novel mechanism of innate immune sensing that may aid in distinguishing pathogenic microbes from non-pathogens . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"infectious",
"diseases/bacterial",
"infections",
"immunology/immune",
"response",
"immunology/innate",
"immunity"
] | 2011 | Secreted Bacterial Effectors That Inhibit Host Protein Synthesis Are Critical for Induction of the Innate Immune Response to Virulent Legionella pneumophila |
Lignin is incorporated into plant cell walls to maintain plant architecture and to ensure long-distance water transport . Lignin composition affects the industrial value of plant material for forage , wood and paper production , and biofuel technologies . Industrial demands have resulted in an increase in the use of genetic engineering to modify lignified plant cell wall composition . However , the interaction of the resulting plants with the environment must be analyzed carefully to ensure that there are no undesirable side effects of lignin modification . We show here that Arabidopsis thaliana mutants with impaired 5-hydroxyguaiacyl O-methyltransferase ( known as caffeate O-methyltransferase; COMT ) function were more susceptible to various bacterial and fungal pathogens . Unexpectedly , asexual sporulation of the downy mildew pathogen , Hyaloperonospora arabidopsidis , was impaired on these mutants . Enhanced resistance to downy mildew was not correlated with increased plant defense responses in comt1 mutants but coincided with a higher frequency of oomycete sexual reproduction within mutant tissues . Comt1 mutants but not wild-type Arabidopsis accumulated soluble 2-O-5-hydroxyferuloyl-l-malate . The compound weakened mycelium vigor and promoted sexual oomycete reproduction when applied to a homothallic oomycete in vitro . These findings suggested that the accumulation of 2-O-5-hydroxyferuloyl-l-malate accounted for the observed comt1 mutant phenotypes during the interaction with H . arabidopsidis . Taken together , our study shows that an artificial downregulation of COMT can drastically alter the interaction of a plant with the biotic environment .
Lignin accounts for about 30% of the organic carbon in the biosphere and is the second most abundant terrestrial biopolymer after cellulose [1] . Lignin fortifies plant cell walls , is the essential composite of wood , and allows terrestrial plants to gain size and volume . This polymer renders the walls of water-conducting cells impermeable , making it possible for the xylem vessels to transport water [2] . Plants with high lignin content are less accessible to microbes and insect herbivores , and the incorporation of this polymer into cell walls is an important mechanism of plant defense against pathogen attack [3] . Lignin is a complex aromatic heteropolymer composed of the three phenylalanine-derived monolignols , p-coumaryl- , coniferyl- , and sinapyl alcohol [1] , [4] ( Figure 1 ) . Polymers composed of these monolignols are referred to as p-hydroxyphenyl ( H ) , guaiacyl ( G ) , and syringyl ( S ) lignin , respectively . Dicotyledonous plants contain mostly G and S lignin , whereas monocotyledonous plants contain H , G , and S lignin [5] . The amount of lignin and the proportions of the H , G , and S subunits within the lignocellulosic biomass determine the technological value of the raw plant material for forage , biofuel distillation , wood , or paper production . Attempts to modify lignin composition have been a major focus of research for decades , with the first mutant described as early as 1935 [6] . The brown midrib-3 ( bm3 ) maize mutant has a lower S-lignin content than normal maize [7] , due to the downregulation of caffeate O-methyltransferase ( COMT ) activity [8] ( Figure 1 ) . COMT has since been targeted repeatedly in efforts to improve the technological quality of plant material by transgenic approaches . The downregulation of COMT activity leads to a decrease in the proportion of S-lignin , with no major change in overall lignin content [9] . As previously reported for bm3 maize [10] , an increase in the G/S lignin ratio generally seems to improve the digestibility of transgenic forage [11] . Poplar trees in which COMT expression has been silenced have been shown to generate higher pulp yields , although the lignins they contain are less amenable to industrial degradation [12] . In field trials , the transgenic trees grew normally and displayed no obvious physiological alterations [13] . However , studies investigating the side effects of changes in COMT gene expression in mutant and transgenic plants remain scarce , although it was shown 30 years ago that the bm3 mutation decreases vascular integrity and stalk robustness in maize [14] . Even less is known about the outcome of biotic interactions involving plants with genetically modified COMT expression and S-lignin content . Arabidopsis thaliana appears to be an excellent tool for addressing this issue . This plant has a typical dicotyledonous lignification pattern [15] , the COMT gene has been clearly identified [16] , tools for functional analyses have been generated , and A . thaliana has been established as a host for a large number of microorganisms . In this study , we used wild-type A . thaliana and mutant A . thaliana lines with inactivated COMT1 for analyzing their phenotypes of interaction with the bacterial pathogens Xanthomonas campestris pv . campestris and Pseudomonas syringae pv . tomato , the fungal pathogens Alternaria brassicicola , Blumeria graminis , and Botrytis cinerea , the oomycete Hyaloperonospora arabidopsidis ( formerly H . parasitica ) , and the root-knot nematode Meloidogyne incognita . The comt1a mutant was found to have much lower levels of S-lignin than the wild type , although its overall lignin content was similar [15] . Knockout of comt1 led to a strong decrease in soluble 2-O-sinapoyl-L-malate ( sinapoyl malate ) levels and the accumulation of 2-O-5-hydroxyferuloyl-L-malate ( hydroxyferuloyl malate ) , which is not detectable in wild-type plants ( [4] , [15] and this study ) . We paid particular attention to interactions with the oomycete H . arabidopsidis , because comt1 mutants displayed an unexpected phenotype when inoculated with this pathogen . Oomycetes have devastating effects on crops , forests , and natural ecosystems , and there are currently no efficient methods for their control [17] . These fungus-like pathogens have unique physiological characteristics and can undergo vegetative or sexual reproduction . Asexual reproduction leads to the creation of clonal populations well adapted to a given host and environment . Sexual reproduction can occur in a single strain , when the species is homothallic . In heterothallic species , two strains of opposite mating types are required for fertilization . In both homothallic and heterothallic species , fertilization results in thick-walled zygotes called oospores , which are produced in infected plant tissue , and released into the soil as the plant tissue degrades . Oospores are highly resistant against environmental influences and can persist in the soil for several years [18] . Sexual reproduction is frequently initiated in response to selective pressure from the environment [19] . The findings presented here indicate that mutations in COMT1 alter the plant physiology in a way that stimulates the oomycete to reproduce sexually . Risk assessments for plants with modified lignin biosynthesis should , therefore , go for quantitative studies of the interaction with the biotic environment , and beyond .
Two allelic mutant lines for COMT1 in the Wassilewskija ( WS ) background were identified from the Versailles Arabidopsis insertion mutant collection [20] . The comt1a and comt1b mutants were homozygous for unique T-DNA insertions in the first exon and the first intron of the At5g54160 locus , respectively ( Figure S1A ) . Comt1a has been shown to be a positive promoter-trap line expressing a functional ß-glucuronidase ( GUS ) gene under the control of the COMT1 promoter [15] . Comt1b harbors the GUS gene as an in-frame insertion in the first intron ( Figure S1A ) , providing GUS activity upon transcriptional activation of COMT1 . For mutant phenotype complementation , a poplar PtOMT1 cDNA was placed under the control of a constitutive promoter and transferred into the comt1a mutant . PtOMT1 was used to avoid the silencing effects observed in Arabidopsis when the COMT1 gene is overexpressed [15] . The complementation line , CpOMT14 , had almost normal levels of COMT activity and S-lignin [15] . The exon-tagged comt1a mutant lacked the COMT1 mRNA signal found in wild-type and CpOMT14 plants on reverse transcription-PCR analysis ( RT-PCR , Figure S1B ) . The T-DNA insertion in comt1b did not lead to a full gene knock-out , but did decrease levels of COMT1 mRNA ( Figure S1B ) , which was correctly spliced and had a sequence identical to that of the wild-type COMT1 mRNA ( data not shown ) . The expression profile of COMT1 was analyzed by GUS staining plant tissues from the promoter-trap lines comt1a and comt1b . Identical profiles of GUS activity were detected in tissues from both lines ( data not shown ) . Under normal growth conditions , GUS activity was restricted to tissues undergoing lignification , such as the vascular systems within aerial organs ( Figure 2A–D ) and the root central cylinder ( Figure 2J ) . Additional constitutive expression was observed within the distal focus [21] of cotyledons ( Figure 2C ) and developing young leaves , within leaf primordia ( Figure 2D ) and root cap columella cells ( Figure 2I ) , suggesting that COMT1 expression responds to venation patterning and auxin signaling [22] . Upon infection with the biotrophic oomycete leaf pathogen , H . arabidopsidis , COMT1 expression extended to mesophyll parenchyma in leaves ( Figure 2E and 2F ) , cotyledons ( Figure 2G ) , and the hypocotyl ( Figure 2H ) . Expression was restricted to cells close to invading hyphae ( Figure 2F ) . In roots , the biotrophic root-knot nematode , M . incognita , invades cortex cells , migrates to the root tip , enters the central cylinder , and moves upward before settling and inducing the formation of a feeding site [23] . This nematode activated COMT1 expression early in cells of the swelling gall ( Figure 2K ) , in giant cells and surrounding dividing cells . COMT1 expression was maintained in the mature root gall until 21 days after inoculation ( Figure 2L ) , in both the giant cells and their neighboring cells . Thus , parenchyma cells from different plant organs did not express COMT1 under normal growth conditions . However , the same cells responded to pathogen infection by local transcriptional activation of this gene . We compared the susceptibility of wild-type plants and of the comt1a mutant to diverse pathogenic microorganisms , including fungi , bacteria , and nematodes ( for details , see Protocol S1 ) . We found the mutant significantly more susceptible to a strain of the necrotrophic fungus , Botrytis cinerea [24] ( Figure S2A ) . Increased susceptibility was also observed for two other analyzed fungal pathogens , the necrotrophic Alternaria brassicicola [25] and the biotrophic Blumeria graminis f . sp . hordei ( Bgh ) [26] ( Figure S2B and S2C ) . In addition , the comt1 mutant was found to be significantly more susceptible to the xylem-colonizing systemic bacterial pathogen Xanthomonas campestris pathovar campestris ( Xcc ) [27] , and the bacterial speck agent , Pseudomonas syringae pv . tomato ( Pst ) ( Figure S3A and S3B ) . The conclusion from these experiments was that a mutation of COMT1 generally weakens plant resistance to microbial pathogens . An exception from this observation was only found for the root-knot nematode M . incognita . Although the M . incognita infection activated COMT1 ( Figure 2 ) , a gene knockout did not influence the mean number of galls established three weeks after inoculation . It did neither affect nematode ability to complete its life cycle , which was characterized by similar amounts of egg masses produced by M . incognita two months after inoculation in mutant and wild-type plants ( Figure S2D ) . Transcriptional activation of COMT1 occurred as a consequence of oomycete infection ( Figure 2 ) . We therefore analyzed whether an inactivation of COMT1 influences the interaction between H . arabidopsidis and A . thaliana under laboratory conditions . The virulent isolate Emwa1 completed its infection cycle on wild-type and comt1 mutant plants , resulting in asexual reproduction and the formation of conidia . However , sporulation levels were 40 to 50% lower on comt1 mutants than on wild-type plants ( Figure 3A ) . This enhanced resistance phenotype of seedlings was partly complemented by expressing PtOMT1 in the comt1a mutant background ( Figure 3A ) , and was confirmed on adult plants , with true leaves from different rosette stages displaying significantly lower levels of H . arabidopsidis sporulation in the absence of COMT1 ( Figure 3B ) . We investigated whether the observed phenotype of comt1 mutants resulted from a gain-of-function in defense signaling , by assessing constitutive and inducible responses before and after inoculation with H . arabidopsidis . The markers used for the activation of salicylic acid ( SA ) - and jasmonic acid ( JA ) -dependent responses were the transcriptional activation of PR-1a and PDF1 . 2b , respectively . In quantitative RT-PCR experiments , PR-1a mRNA was undetectable in untreated wild-type and mutant plants . PDF1 . 2b transcripts were present at similar , low levels in both lines tested in the absence of inoculation . The accumulation of transcripts from both genes was induced by the pathogen , to similar levels in wild-type and mutant plants ( Figure S4 ) . The enhanced resistance phenotype of comt mutants thus appeared not to be dependent on SA and JA-dependent defense signaling pathways in Arabidopsis . Following the inoculation of wild-type A . thaliana , H . arabidopsidis spores germinate on leaf surfaces and form appressoria , enabling the penetration pegs to overcome the cuticle . Once inside the leaf , the hyphae grow intercellular , branch , and establish a filamentous network spanning the entire infected leaf three to four days after the onset of infection ( Figure 4A ) . The growing hyphae locally digest the cell walls of almost all the plant cells they come into contact with , leading to invagination of the host plasma membrane and the formation of intracellular bulbous structures called haustoria ( Figure 4D ) . Haustoria are feeding sites required for the biotrophic lifestyle of the oomycete . Four to six days after inoculation under laboratory conditions , the hyphae use stomatal openings to form conidiophores on the leaf surface and initiate asexual reproduction [28] . In comt1 mutant lines , the initial infection process was identical to that in wild-type plants . H . arabidopsidis penetrated , formed a filamentous network of branched hyphae ( Figure 4B ) , and developed haustoria within host cells . We quantified the development of the oomycete in wild-type and mutant Arabidopsis , by performing RT-PCR with gene-specific primers to generate an amplicon within the intergenic transcribed spacer ( ITS2 ) of H . arabidopsidis ribosomal RNA ( Figure 4C ) . These experiments showed that the oomycete expanded similarly in wild-type and mutant plants ( Figure 4C ) , indicating that hyphal growth and branching were not affected in the mutants . However , microscopic analyses of haustoria indicated that the structure of the feeding sites was less stable in comt1 mutant plant cells than in wild-type cells . Disintegration of the haustorium was observed in mutant cells ( Figure 4E ) but never in the wild-type background . Even more strikingly , the frequency of sexual reproduction was found to be higher in comt1 mutants , with a significantly larger number of oospores within mutant tissues than within wild-type tissues ( Figure 4F ) . This phenotype was almost complemented in the CpOMT14 line ( Figure 4F ) . Thus , the lower level of conidiospore formation in comt1 mutants ( Figure 3 ) coincided with a higher frequency of sexual reproduction . Plant cells store excess monolignol precursors not incorporated into lignin as soluble esters . In the leaves and cotyledons of A . thaliana , the major accumulating soluble hydroxycinnamate ester is sinapoyl malate ( SM ) [29] . In the comt1a mutant , SM levels are much lower , with these mutants instead accumulating hydroxyferuloyl malate ( OH-FM ) , which is a derivative of the COMT1 substrate , 5-hydroxyconiferaldehyde [4] , [15] . We quantified the accumulation of hydroxycinnamoyl malate esters in the various lines used in this study and evaluated the biological activities of these compounds , by synthesizing SM , OH-FM , and 2-O-feruloyl-L-malate ( feruloyl malate; FM ) ( see Protocol S2 , and Figure S5 ) . On reverse-phase high-performance liquid chromatography ( HPLC ) of methanolic extracts from four-week-old plantlets , FM was not detected in any of the Arabidopsis lines . Wild-type plants accumulated soluble SM to a concentration of about 280 nmol/g fresh weight and OH-FM was not detectable in these plants ( Figure 5A ) . By contrast , the mutant lines comt1a and comt1b accumulated OH-FM to concentrations of 200 nmol and 150 nmol/g fresh weight , respectively , whereas SM levels were about 70% lower than wild-type levels in comt1a , and 55% lower in comt1b ( Figure 5A and Figure S3 ) . SM was again the main compound detected in seedlings from the complemented mutant line CpOMT14 ( Figure 5A ) . The COMT1 mutation modified the SM/OH-FM ratio , but did not modify total hydroxycinnamoyl malate ester concentrations , which appeared to be similar in plants from all lines . The correlation between the accumulation of soluble OH-FM ( Figure 5A ) and the frequency of sexual reproduction ( Figure 4F ) in the comt1 mutant lines led us to analyze whether this compound was able to stimulate oospore formation . We thus developed an in vitro assay for the sexual reproduction of another homothallic oomycete plant pathogen , Phytophthora cactorum . Using P . cactorum zoopores , we synchronized the age and density of hyphae in titer plate wells before adding OH-FM , SM , or FM at various concentrations . In the absence of these compounds , P . cactorum produced about 5 , 000 oospores ( per g fresh weight of mycelium ) . The addition of SM or FM did not alter the frequency of sexual reproduction of the oomycete ( Figure 5B ) . In contrast , supplementation of the medium with OH-FM significantly stimulated sexual reproduction , and the number of oospores doubled when OH-FM was added to concentrations of about 0 . 1 µM ( Figure 5B ) . Beyond 0 . 1 mM , the effect of OH-FM on sexual reproduction was less obvious . At these concentrations , OH-FM appeared to influence the integrity of the hyphal network , and the mycelium developing in the presence of the compound became frail . These findings strongly suggest that the accumulation of OH-FM can account for the greater sexual activity of H . arabidopsidis in comt1 mutants . However , OH-FM did not stimulate the sexual reproduction of individual mating types of the heterothallic oomycete , P . parasitica ( data not shown ) . OH-FM therefore cannot replace an oomycete mating hormone .
In recent decades , lignin has become a privileged target for genetic engineering approaches aiming to increase the industrial processing efficiency of plant biomass . However , the pros and cons of the use of such transgenic plants in open-field situations have only recently begun to be explored [30]–[32] . Systems biology approaches have been used to investigate the extent to which a single engineered gene encoding a protein involved in the lignin biosynthesis pathway interferes with other metabolic pathways , and the extent to which it modifies interactions between the modified plant and its environment [33] . The downregulation of CAD and CCR ( Figure 1 ) in tobacco has been shown to affect not only lignin biosynthesis , but also primary metabolism , stress metabolism , and photorespiration [34] . In poplar , the downregulation of CCR affects the overall metabolism and structure of cell wall polymers [35] . This enzyme also directly regulates pathogen defense signaling in rice [3] , indicating that lignin biosynthesis plays a more subtle role in plant defense responses against herbivores [36] and microbes [37] than simply constituting a physical barrier . This report provides the first large-scale analysis of the role of COMT in interactions between the plant and its biotic environment . We found that the transcriptional activation of COMT1 in parenchyma tissues from roots and aerial organs of Arabidopsis was stimulated by pathogen infection . COMT1 expression was shown before to be enhanced upon Arabidopsis leaf infiltration with either the nonhost bacterium P . syringae pv . phaseolicola [38] , bacterial flagellin ( flg22 ) [38] , [39] , or harpin ( HrpZ ) [38] , [40] , or with the necrosis-inducing Phytophthora protein 1 ( NPP1 ) [38] , [41] . The transcription of COMT1 appears thus to correlate with the onset of the plant's pathogen-associated molecular pattern ( PAMP ) -triggered immune ( PTI ) response [42] . In the Arabidopsis leaf mesophyll , we found an accumulation of toluidine blue-stainable , lignin-like deposits in host cells that surround H . arabidopsidis hyphae ( data not shown ) . Taken together , we suppose that the observed COMT1 promoter activation in response to infection reflects an onset of PTI within host cells and an attempt to reinforce wall rigidity through lignification . Consequently , we found that COMT1 downregulation increased susceptibility to at least three fungal and two bacterial pathogens . Mutants were significantly more susceptible to a moderately aggressive strain of the necrotrophic fungus , B . cinerea , but not to a highly virulent isolate . These findings indicate that COMT1 contributes to limiting the spread of manageable strains of the fungus . The comt1a mutant was also more susceptible to another necrotrophic fungal pathogen , Alternaria brassicicola . This fungus causes black spot disease on almost all cultivated Brassica species including broccoli , cabbage , canola and mustard . It is of worldwide economic importance , because it reduces crop yields and the quality of canola oil [43] . From a human health perspective , Alternaria brassicicola belongs to a genus of fungi considered one of the most potent sources of mold-derived allergens [44] . The resistance of A . thaliana to this fungus requires the phytoalexin camalexin and the signaling molecule JA , but is independent of SA [25] . The downregulation of COMT1 weakened the JA-dependent defenses of Arabidopsis against this pathogen . Unlike B . cinerea and Alternaria brassicicola , Bgh is a biotrophic fungal pathogen causing powdery mildew on barley . Arabidopsis is a nonhost for this pathogen , and SA-dependent defense responses are activated rapidly when the pathogen attempts to penetrate [26] . The higher proportion of successful infections in the mutant indicated that the mutation of COMT1 weakens this penetration resistance of A . thaliana . Mutants were also more susceptible to the bacterial pathogens Pst carrying avrPphB , and Xcc . AvrPphB is recognized by the corresponding resistance gene product , RPS5 , which is present in the WS background , but absent in the Ler background [45] . Plants of these two ecotypes are thus resistant and susceptible , respectively , to Pst avrPphB . The observation that COMT1 downregulation renders WS as susceptible as Ler indicates that the gene plays also an important role in avrPphB-mediated resistance to the bacterial pathogen . Two exceptions to the trend of comt1a mutants being generally more susceptible to pathogens were observed after inoculation with the root-knot nematode , M . incognita , and the biotrophic oomycete pathogen , H . arabidopsidis . M . incognita triggers the transcriptional activation of COMT1 , but knocking out this gene had no effect on disease establishment ( gall formation ) , or nematode development ( egg masses ) . However , all the Arabidopsis ecotypes analyzed to date are susceptible to M . incognita . Variety-specific differences in defense and resistance influencing nematode development , such as those within the Solanaceae [46] , are not known for A . thaliana . We therefore cannot exclude the possibility that COMT1 may affect resistance phenotypes in other plant species . In this context , it has been reported that a downregulation of the COMT1 ortholog in tobacco [47] , leads to increased M . incognita reproduction [48] . The obligate oomycete pathogen H . parasitica causes downy mildew disease on agronomically important Brassicaceae , such as rapeseed and cabbage . Its relative , H . arabidopsidis only infects A . thaliana in clear gene-for-gene relationships and is nonpathogenic on other crucifers tested [49]–[51] . WS harbors genes from the RPP1 group ( RPP1-WsA , RPP1-WsB , and RPP1-WsC ) , which confer resistance to the H . arabidopsidis isolate Noco2 , but not to Emwa1 [50] . These two isolates thus give rise to genetically incompatible and compatible interactions , respectively , with WS . In Arabidopsis seedlings inoculated with Noco2 , comt1 mutants displayed the same resistance phenotype as WS , with no hyphal development detectable in any of the lines tested ( data not shown ) . COMT1 is thus not a key enzyme for the genetic resistance of Arabidopsis to oomycetes . However , the virulent H . arabidopsidis isolate Emwa1 produced significantly fewer conidiospores on COMT1 mutants than on wild-type plants . According to the generally accepted criterion for analyzing plant susceptibility to this oomycete [52] , comt1 mutants were thus more resistant . To date , several Arabidopsis genes have been identified , which are required for full susceptibility to biotrophic fungal and oomycete pathogens , and which confer , when inactivated , increased resistance phenotypes to mutants . Vogel and coworkers identified 26 recessive powdery mildew resistance ( pmr ) mutants in a genetic screen [53] . PMR6 codes for a pectate lyase [54] , and PMR2 is the Arabidopsis ortholog of the barley mlo gene , Atmlo2 [55] . The MLO protein is required for successful entry of Bgh into the host cell [56] . Both pmr6 and Atmlo2 mutants were not affected in susceptibility to H . arabidopsidis [54] , [55] , indicating that the downy mildew does not require the host cell pectate lyase and MLO for successful infection . However , another pmr mutant , pmr4 ( or gsl5 ) is more resistant to both powdery- and downy mildews [53] . PMR4 codes for a callose synthase , which seems to be involved in yet unknown functions required for functional haustoria formation [57] , and which negatively regulates SA defense signaling in the plant [58] . Since , six additional Arabidopsis loci were identified that confer , when mutated , resistance to H . arabidopsidis [59] . For three of the downy mildew resistant ( dmr ) mutants , enhanced resistance was correlated with a constitutive expression of the SA-dependent PR-1a gene [59] . Of the remaining DMR genes , DMR6 was identified to code for a 2-oxoglutarate ( 2OG ) -Fe ( II ) oxygenase of unknown function [60] . The role of DMR6 for disease susceptibility is not yet known , but it also appears to be a negative regulator of defense gene activation [60] . Here , we showed that the increased resistance of comt1 mutants was not correlated with a change in PR-1a and PDF1 . 2 gene expression , when compared to wild-type plants . The inactivation of COMT1 appears , therefore , not to interfere with SA- and JA-dependent defense signaling pathways . The absence of COMT1 did not impair intercellular hyphal growth and branching , and did not interfere with overall oomycete biomass development in mutant plant tissues . However , H . arabidopsidis interacting with mutant plants displayed haustorial instability and enhanced sexual reproduction as two additional phenotypes to reduced asexual sporulation . We found that the occurrence of these phenotypes correlated with a metabolic difference between comt1 mutants and wild-type plants , i . e . lower levels of SM within mutants and accumulation of OH-FM instead . Synthetic OH-FM promoted the sexual reproduction of P . cactorum in vitro , whereas the closely related derivative SM , which was present in large amounts in wild-type Arabidopsis , and synthetic FM did not . Moreover , high OH-FM concentrations decreased the mechanical resistance of the P . cactorum hyphal network in vitro , indicating that the compound is toxic for oomycetes and likely responsible for the observed haustorial instability phenotype of H . arabidopsidis in comt1 mutants . These findings make OH-FM accumulation a potential candidate cause for the phenotypes we observed during the interaction between comt1 mutants and H . arabidopsidis . In this context , it might be possible that the observed transcriptional activation of COMT1 in host cells close to downy mildew hyphae has another reason than the above discussed stimulated lignification defense . In a hypothetical scenario , an H . arabidopsidis virulence function might promote host COMT1 expression , in order to prevent plant cells harboring oomycete haustoria from any accumulation of detrimental OH-FM . However , OH-FM was not detectable in wild-type Arabidopsis by the means we used for the present study . To prove the hypothetical scenario , more sensitive detection methods for OH-FM need to be developed to compare the accumulation of microquantities of the compound in living cells harboring or not haustoria . In conclusion , this study demonstrates that manipulating the expression of single genes within the monolignol biosynthetic pathway may affect the interaction of engineered plants with the biotic environment . COMT has been a key target for such manipulation in the past [13] . The downregulation of COMT in agronomically important plants may affect susceptibility to fungal and bacterial attack in a similar manner , as shown here for A . thaliana . The metabolic disequilibrium generated by COMT1 knockout in Arabidopsis probably also creates a selective pressure in other crops that forces oomycete pathogens to undergo sexual reproduction . Sexual reproduction is a source of genetic variation even in homothallic oomycetes [61] , [62] . Furthermore , oospores persist in the soil [63] , [64] and are insensitive to fungicides [65] , and thus make disease control difficult . The stimulation of oomycete sexual reproduction in genetically engineered COMT plants may , therefore , lead to the evolution of novel infection traits .
All Arabidopsis lines used for the experiments were from the Wassilewskija ( WS ) genetic background . The comt1a mutant , previously named Atomt1 [15] and comt1 [4] , and the complemented line CpOMT14 have been described before [15] . The comt1b mutant from the Versailles T-DNA insertion collection [18] was obtained from Dr . Laurent Nussaume ( CEA , Cadarache , France ) . Analysis of the T-DNA flanking region within comt1b was performed by sequencing the amplicon obtained with primer pair 3 and 4 on genomic DNA as the template ( Figure S1 ) . Plants were grown in sand supplemented with MS medium in growth chambers at 20°C with a 12 h photoperiod . H . arabidopsidis isolates Emwa1 and Noco2 were obtained from Dr . Jane Parker ( MPIZ , Cologne , Germany ) , and transferred weekly onto the genetically susceptible Arabidopsis accessions WS and Columbia ( Col ) , respectively , as described [66] . For infection , 10-day-old plants were spray-inoculated to saturation with a spore suspension of 40 , 000 spores/ml . Plants were kept in a growth cabinet at 16°C for 3 d with a 16 h photoperiod . Sporulation was then induced by spraying plants with water , and keeping them for 48 hours under high humidity . To evaluate conidiospore production , pools of 8 plants were harvested in 1 ml of water . After vortexing , the amount of liberated spores was determined with a hemocytometer . To evaluate oospore production , inoculated cotyledons were destained in 80% ethanol , mounted on glass slides , and analyzed by microscopy . For nematode infection , A . thaliana were grown in vitro on MS medium containing 1% sucrose and 0 . 7% plant cell culture-tested agar ( Sigma-Aldrich ) . One hundred surface-sterilized freshly hatched M . incognita J2 larvae were added to each 2-week-old seedling , as described [67] . The plates were kept at 20°C with a 16 h photoperiod . Pathogenicity assays with B . cinerea , A . brassicicola , B . graminis , P . syringae , and X . campestris were performed as described in Protocol S1 . GUS activity in comt1a mutants was analyzed histochemically [68] . For the observation of oomycete development in infected tissues , cotyledons were fixed in 0 . 1 M glutaraldehyde and 1 . 5 M formaldehyde in phosphate buffer , bleached in a series of increasing ethanol concentrations , and stained with 0 . 3% Fluorescent Brightener 28 [69] . Calcofluor-stained hyphae were observed by fluorescence binocular microscopy , or a Zeiss Axioplan 2 fluorescence microscope configured for brightfield , darkfield , and differential interference contrast ( excitation 480 nm , barrier filter 510 nm ) . Total RNA was extracted from A . thaliana seedlings using TRIZOL Reagent ( Invitrogen ) following the instructions of the manufacturer . One µg RNA was reverse-transcribed using the iScript cDNA Synthesis Kit ( Biorad ) . For H . arabidopsidis ITS2 RNA quantification within infected tissues , PCR ( 25 cycles ) was performed with 8 ng cDNA as template using the ITS2 forward primer 5′-TGTGGTAGACGAATGGGTGA-3′ , and the ITS2 reverse primer 5′-AAGTGCAGCCGAAGCTTTAC-3′ . PCR with the primer pairs OXA1-a and OXA1-b ( Figure S1 ) was used as a quantitative control . Aliquots of individual PCR products were resolved by agarose gel electrophoresis , visualized with ethidium bromide and quantified using a Fujifilm FLA-3000 Phospho/Fluoroimager . The expression of PR-1a and PDF1 . 2b was analyzed by real-time quantitative PCR using the forward primer 5′-GGAGCTACGCAGAACAACTAAGA-3′ and reverse primer 5′-CCCACGAGGATCATAGTTGCAACTGA-3′ for PR-1a , and the forward primer 5′-TCATGGCTAAGTTTGCTTCC-3′ , and reverse primer 5′-AATACACACCACGATTTAGCACC-3′ for PDF1 . 2b . Amplification and detection were performed in the Chromo4 detection system ( Biorad ) . Reactions were done in a final volume of 15 µl containing 10 µl qPCR MasterMix Plus For SYBRGreen I No Rox ( Eurogentec ) , 0 . 5 mM of each primer , and 8 ng of cDNA template . PCR conditions were as follows: 95°C for 15 min , followed by 40 cycles of 95°C for 15 s , 56°C for 30 s and 72°C for 30 s . At the end of the program a melting curve ( from 60°C to 95°C , read every 0 . 5°C ) was determined to ensure that only single products were formed . Ubiquitin-specific protease 22 ( UBP22 ) expression was used to normalize the transcript level in each sample with the primer pairs 5′-GCCAAAGCTGTGGAGAAAAG-3′ and 5′-TGTTTAGGCGGAACGGATAC-3′ . Data analysis was performed using the MJ OpticonMonitor Analysis software ( version 3 . 1; Biorad ) . Extraction of soluble phenolics from fresh plants was carried out in 100% methanol ( 0 . 4 ml/100 mg of fresh weight ) . After centrifugation , the cleared supernatant was adjusted to 80% aequeous methanol , and passed through a 200 mg C18 ( Nucleodur 100-30 , Macherey-Nagel , Düren , Germany ) , in order to stop chorophyll and lipidic compounds . The unretained fraction was directly analyzed by HPLC . The liquid chromatograph ( System Controller 680 with two 510 pumps; Waters Millipore , Milford , MA ) was equipped with an Inertsil 5ODS3 C18 column ( 5 µm , 250×4 , 6 mm i . d . ; Interchim , Montluçon , France ) . Samples were chromatographied with the following gradient ( flow rate l ml/min ) : 1 min isocratic 15% solvent A ( methanol ) and 85% solvent B ( water , 0 . 5% H3PO4 ) , then within 29 min to 75% solvent A , then within 3 additional min to 100% solvent A , followed by a 3 min isocratic step in 100% solvent A . Compounds were detected by a Waters TM 996 Photodiode Array Detector ( 200 to 500 nm ) . Peaks were identified and quantified using the Empower software ( Waters ) , after external standardization with synthetic compounds . The P . cactorum isolate 723 was from the Sophia Antipolis Phytophthora collection , and was sampled in 2005 from Fragaria in France . Propagation and zoospore production were performed as described [70] , and 5 , 000 zoospores were pipetted into 0 . 5 ml of V8 medium [70] in 24-well titer plates . After 24 h at 24°C , another 0 . 5 ml of V8 medium containing or not the hydroxycinnamoyl malates were added . After an incubation for further 48 h at 24°C , mycelium was picked out of the wells , rinsed , dried on filter paper , weighed , placed into 1 ml of water , and macerated in a potter . Liberated oospores were enumerated with a hemocytometer and the number of oospores was expressed per g mycelium fresh weight . Sequences used in this article were derived from gene IDs 835504 ( COMT1 ) , 169452 ( PtOMT1 ) , 1815949 ( PR1-a ) , 817143 ( PDF1 . 2b ) , 836325 ( OXA1 ) , 830946 ( UBP22 ) . | Lignin is an essential component of wood and the second most abundant terrestrial biopolymer . Plants synthesize lignin to strengthen cell walls and to resist pathogen attack . Because the technological value of plants is determined by the amount and composition of lignin , genetic engineering approaches frequently aim at altering these parameters . However , data showing how plants with modified lignin content interact with the biotic environment are still scarce . This study is the first of its kind to evaluate how a model plant , which was mutated for a key enzyme in lignin biosynthesis , interacts with pathogens from different kingdoms . We found that the mutants were generally more susceptible to bacteria and fungi . Additionally , the mutation altered the plant physiology in a manner that elevated sexual reproduction of an oomycete pathogen . Oospores resulting from this mode of reproduction are the most persisting propagules of this pathogen . Their elevated formation correlated with the accumulation of a soluble phenylpropanoid derivative in the mutants , which we synthesized and found to be able to stimulate sexual oomycete reproduction in vitro . Our study indicates that interfering with lignin composition may drastically alter the outcome of plant–pathogen interactions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"chemical",
"biology/small",
"molecule",
"chemistry",
"microbiology/environmental",
"microbiology",
"microbiology/plant-biotic",
"interactions",
"plant",
"biology/plant-biotic",
"interactions",
"plant",
"biology/plant",
"biochemistry",
"and",
"physiology",
"plant",
"biology/agricultural",
"biotechnology"
] | 2009 | Imbalanced Lignin Biosynthesis Promotes the Sexual Reproduction of Homothallic Oomycete Pathogens |
Amino-acid coevolution can be referred to mutational compensatory patterns preserving the function of a protein . Viral envelope glycoproteins , which mediate entry of enveloped viruses into their host cells , are shaped by coevolution signals that confer to viruses the plasticity to evade neutralizing antibodies without altering viral entry mechanisms . The functions and structures of the two envelope glycoproteins of the Hepatitis C Virus ( HCV ) , E1 and E2 , are poorly described . Especially , how these two proteins mediate the HCV fusion process between the viral and the cell membrane remains elusive . Here , as a proof of concept , we aimed to take advantage of an original coevolution method recently developed to shed light on the HCV fusion mechanism . When first applied to the well-characterized Dengue Virus ( DENV ) envelope glycoproteins , coevolution analysis was able to predict important structural features and rearrangements of these viral protein complexes . When applied to HCV E1E2 , computational coevolution analysis predicted that E1 and E2 refold interdependently during fusion through rearrangements of the E2 Back Layer ( BL ) . Consistently , a soluble BL-derived polypeptide inhibited HCV infection of hepatoma cell lines , primary human hepatocytes and humanized liver mice . We showed that this polypeptide specifically inhibited HCV fusogenic rearrangements , hence supporting the critical role of this domain during HCV fusion . By combining coevolution analysis and in vitro assays , we also uncovered functionally-significant coevolving signals between E1 and E2 BL/Stem regions that govern HCV fusion , demonstrating the accuracy of our coevolution predictions . Altogether , our work shed light on important structural features of the HCV fusion mechanism and contributes to advance our functional understanding of this process . This study also provides an important proof of concept that coevolution can be employed to explore viral protein mediated-processes , and can guide the development of innovative translational strategies against challenging human-tropic viruses .
Flaviviridae such as Hepatitis C Virus ( HCV ) , Dengue Virus ( DENV ) , Zika Virus ( ZIKV ) or West Nile Virus ( WNV ) are cause of several acute and chronic diseases worldwide . The continuous investigation of the molecular processes by which these RNA viruses infect and replicate into their host is critical to develop innovative anti-viral strategies and anticipate viral resistance to pre-existing drugs . The high potency of viral genomes to mutate is often considered as a major limitation for the development of effective anti-viral strategies . Nevertheless , the high-mutation rate of RNA viruses represents a unique opportunity to decrypt viral protein functions and structures . Highly-evolving viral genomes are shaped by important evolutionary constraints to maintain genetic structure and proper protein folding . Amino-acid coevolution , which refers to mutations of different residues at a similar time frame , mirrors such constraints . Hence , the identification and characterization of coevolution signals imprinted within viral protein sequences can provide unique insights into viral protein functions and conformational changes , and ultimately guide the design of original anti-viral strategies . Virus entry is a conserved , critical step during the viral life cycle and represents a valuable target for the development of antivirals and vaccines . The entry process of HCV is orchestrated by two envelope glycoproteins , E1 and E2 , which are incorporated onto the virion surface . During entry , E1E2 mediate viral particle attachment to cell surface receptors and induce the merging ( called fusion ) of endosomal and virus membranes at acidic pH , thus leading to release of viral RNA into the cytosol [1] . Flaviviruses such as DENV , ZIKV or WNV harbor two envelope glycoproteins: E and PrM . E is a class-II fusion protein composed of three distinct domains ( domain I , II and III; or DI , DII and DIII respectively ) and carries both binding and membrane fusion properties [2] . Crystal structure of E at different pH allowed to draw a fusion model during which initial E dimers change conformation and fold back as trimer structures to induce membrane merging [3] . In contrast , how HCV E1 and E2 mediate membrane fusion remains poorly defined . Our understanding of the HCV fusion process is strongly dampened by the absence of a well-defined pre- and post-fusion full-length E1E2 crystal structure . Few studies have attempted to computationally model pre-fusion E1E2 complex [4 , 5] but their impact remain limited as they have to rely on partial structural and functional information that are often collected in a non-heterodimer context . The structure of a large region of the E2 ectodomain ( E2 core ) [6 , 7] exhibits a globular , non-extended fold divided into two distinct sheets: a front sheet composed of a front layer and a central Ig-fold domain , and a back sheet ( or back layer , BL ) . Although the central Ig-fold domain represents a common structure among class-II fusion proteins , the BL harbors an original structure , thus undermining the possibility that HCV E2 is a classical fusion protein . It has been suggested through the resolution of the bovine viral diarrhea pestivirus ( BVDV-1 ) E2 glycoprotein structure that HCV E1 may represent the HCV fusion protein [8–10] . Several studies have identified a hydrophobic region in E1 ( CSALYVGDLC ) that could represent the putative HCV fusion peptide [11–16] . Another study also suggested that E1 proteins form trimeric structure at the virus surface [17] . However , the recent crystal structure of the N-terminal domain of HCV E1 ectodomain does not harbor the expected truncated class-II fusion protein fold [18] , suggesting that HCV fusion might be a unique process . Mutagenesis studies have shown that both E1 and E2 domains , as well as E1-E2 dialogs , are involved in the HCV fusion process [13 , 14 , 19–21] . Thus , rather than being mediated by a single glycoprotein , HCV fusion appears to be mediated by complex intra- and inter-molecular E1-E2 dialogs that shape structural and conformational rearrangements of the heterodimer complex . Consequently , the characterization of interplays between E1 and E2 is critical to decipher the HCV fusion mechanism . Here , using HCV fusion as a model of study , we aimed to provide a proof of concept that amino-acid coevolution and protein evolutionary constraints can shed light on viral protein functions and rearrangements . We hypothesized that detection of E1-E2 coevolution patterns can uncover their functional interplays as well as critical features of the HCV fusion mechanism . We recently developed an original computational method , Blocks In Sequences ( BIS ) , that can robustly detect coevolution signals within conserved cellular and viral proteins using a limited number of protein sequences [22–24] . Taking advantage of this methodology , we aim at establishing a map of E1-E2 coevolution patterns and test whether coevolution analysis can be employed to gain mechanistic insight into poorly characterized viral processes such as HCV fusion . BIS was able to accurately predict features of DENV glycoproteins structural organization onto viral particle , as well as E fusogenic conformational changes . When applied to HCV E1E2 , BIS suggested that HCV E2 BL is a critical modulator of HCV fusion . Consistently , a soluble form of the E2 BL was able to inhibit HCV fusion . Moreover , coevolution signals between E1 and E2 BL/Stem predicted by BIS were found to regulate virus fusion in vitro . Beyond providing novel insights into the HCV fusion mechanism , our work also demonstrates that coevolution analysis can shed light on viral-mediated processes and can open avenues for the accelerated design of innovative anti-viral compounds against challenging human tropic-viruses .
We have previously reported that BIS , a combinatorial-based coevolution analysis method ( S1 Fig ) , can accurately detect coevolution signals within a wide range of well-characterized cellular and viral proteins [22–24] . As BIS has not been previously tested for its ability to predict viral envelope glycoproteins structural organization and rearrangements , we first performed a coevolution analysis of the well-characterized DENV envelope glycoproteins , E and PrM [2 , 3] . During virus maturation , M protein ( a mainly transmembranous protein ) is associated to Pr , a peptide that protect the E fusion peptide and is cleaved prior viral budding [25] . Briefly , the PrM-E complex protrude as trimer at the surface of immature viral particles in the endoplasmic reticulum , a neutral pH compartment . Immature particle then navigates toward the trans-Golgi network , a more acidic compartment , where PrM-E complex form dimeric structures that lie down onto the surface of the particles . Pr , which initially concealed E fusion peptide , is then cleaved by the Furin . This cleavage achieves the maturation of viral particles that are then released into the extracellular compartment . As M is mostly a transmembrane protein with only very partial structural information available , we aimed to determine whether BIS can recapitulate the diversity of E-Pr structural organization at the surface of immature viral particles . BIS analysis of 17 DENV PrM-E serotype 2 sequences led to the identification of 14 groups of coevolving residues ( possibly organized in blocks of consecutive amino-acids ) , further referred as clusters ( S1 Table ) . Among those , three clusters ( cluster 2 , 7 and 9 ) displayed a strong statistical significance ( with associated p values of 8e-5 , 8e-5 and 7e-3 respectively ) and involved coevolving blocks between E and Pr . When E is assembled as dimer at low pH condition , cluster 2 coevolving block positions supported the close proximity of E DIII and Pr ( Fig 1A and 1B-left ) . Similarly , cluster 7 also recapitulated the close proximity between Pr and E domain DII on a trimeric E-Pr structure that form at neutral pH in the endoplasmic reticulum ( Fig 1A and 1B-center ) . On a trimeric Pr-E structure ( one dimer + one monomer of PrM-E ) found at low pH when particles mature in the trans-Golgi , cluster 9 also supported a close proximity between Pr and E DII , but also between E dimers ( Fig 1A and 1B-right ) . Two other clusters ( cluster 3 and 8 ) of strong statistical significance ( with respective p values of 1 . 34e-5 and 4 . 11e-5 ) were identified by BIS but did not involve coevolving residues located within E protein ( S1 Table ) . Although cluster 8 was composed of two coevolving blocks located within M protein only , clusters 3 involved coevolving residues located within both M and Pr . As cluster 3 , cluster 2 and 9 also supported the existence of interactions between Pr and M ( Fig 1A and 1B ) . As M and Pr are the two cleavage products of a single PrM protein , these three clusters hence represent additional evidence of BIS ability to recapitulate biologically significant protein interactions . Taken together , our results demonstrate that BIS has the ability to accurately predict the tridimensional assembly of two viral proteins within different conformational states . As DENV E has been demonstrated to mediate DENV viral fusion , we then aimed to study the intra-protein coevolution signals within DENV E only ( Fig 2A; see new E numbering by BIS in comparison to Fig 1A ) , and determined whether coevolution signals can also predict E fusogenic rearrangements . Using 17 different DENV E serotype 2 sequences , BIS identified 12 clusters ( S2 Table ) . Among them , nine clusters ( clusters 2–8 and 11 , 12 ) displayed associated p-values ranging between 7e-3 to 2e-7 and two clusters ( clusters 9 and 10 ) exhibited associated p-values of 0 . 058 ( S2 Table ) . Cluster 1 and 2 were either conserved ( p-value = 1 ) or too large respectively to be considered . Several clusters ( 3 , 4 , 6 and 7 ) displayed small blocks located within a single region of E both in the linear protein sequence and on the tridimensional structure , suggesting that coevolution signals might contribute to the structural organization of secondary protein sub-domains ( such as internal loops ) ( S2 Fig ) . DENV E DI and DII are composed of two or three sub-domains that are distant on the linear protein sequence but form single structured domains in the protein tertiary structure . Cluster 8 blocks were mostly located within the two sub-domains of DII and were consistent with the tridimensional organization of this protein domain ( Fig 2B ) . Cluster 8 blocks located within the second sub-domain of DII ( DII-2 ) of each E monomer were in close contact on the dimeric E structure ( especially at the level of the DII α-helix ) , hence suggesting that coevolution signals can predict point of contacts between E monomers once organized as dimer ( Fig 2B ) . Despite lower statistical significance ( p<0 . 06 ) , cluster 9 and 10 coevolving blocks also supported E structural organization as these blocks were distant on the linear structure but close on the E dimer structure ( S2 Fig ) . Finally , three clusters ( 5 , 11 and 12 ) displayed coevolving blocks that were both distant on the linear and tridimensional E structure . During fusion , E DIII folds-over toward DII , and DII becomes at close proximity with the E transmembrane domain [3] ( Fig 2C ) . Cluster 5 and 11 coevolving blocks organization were consistent with these structural rearrangements ( Fig 2C ) , suggesting that BIS can recapitulate viral glycoprotein fusogenic conformational changes . Following validation of BIS ability to model viral envelope glycoprotein structural rearrangements , we then applied the BIS methodology to HCV E1E2 . We analyzed independently using BIS ten groups of E1E2 sequences from different genotypes ( gt ) or sub-types ( 1a , 1b , 1 = 1a+1b , 2a , 2b , 2 = 2a+2b , 3 , 4a , 5a and 6a ) ( S3 Table ) . Interestingly , most of the identified clusters involved residues in both E1 and in E2 suggesting the existence of conserved tight dialogs between E1 and E2 proteins ( Fig 3A ) . Only a few number of statistically significant clusters were found among genotype 4a to 6a sequences , due to the low number of sequences available and to their very low genetic divergence . Unlike Dengue E and PrM , the full panel of HCV E2 conformations still remain unknown and E2core only represent a single of these possible conformations , in a given biochemical context . When BIS coevolution analysis is applied to protein ( s ) for which only a fraction of its/their conformations have been characterized ( which is the case for HCV E1 and E2 ) , BIS coevolution clusters can thus only suggest , but not contradict a given conformational hypothesis , this unless the full panel of a protein conformations is known . Consequently , in vitro and/or in vivo experiments are then critical to ascertain the functional significance of a given conformational hypothesis . Given this particular experimental context and in order to identify putative E1E2 rearrangements during HCV fusion , we thus adopted the following approach . First , we aimed to assign to each E1E2 cluster a given function by mapping clusters blocks with residues previously identified in the literature to impact E1E2 folding/heterodimerization , binding or fusion . Second , we sought to identify among BIS clusters classified as “fusion clusters” a putative protein rearrangement supported by several fusion clusters across multiple genotypes , prior experimental challenge through in vitro experiments . We first focused our efforts on analyzing gt1a clusters . Detailed analysis of these clusters ( S4 Table; S3 Fig; note that BIS numbered E1 and E2 residues by considering the first amino acid of E1 as residue #1 ) showed E1 as coevolving systematically with all the E2 domains . Plotting the gt1a coevolving blocks onto reference sequences ( S4 Fig ) revealed a strong correlation between blocks and residues previously identified in the literature to be important for heterodimer folding or viral binding site conformation ( grouped both under the term “structural” ) or fusion . This correlation allowed us to propose functions ( structural , fusion , or multifunctional clusters ) for most gt1a clusters ( S5 and S6 Tables ) . When plotted on the E2core structure ( Fig 3B ) , most fusion clusters involved blocks located within the E2 BL ( Fig 3C and 3D ) in contrast to structural or multifunctional clusters for which blocks were broadly distributed across E2 ( S5 Fig ) . Interestingly , some of the fusion clusters involved distant blocks in both E1 and E2 ( clusters 5 , 7 , 10; Fig 3C and 3D ) , highlighting that E1 terminal regions and the E2 BL could be in close proximity during fusogenic rearrangements . In addition , BIS also suggested an association between fusogenic rearrangements and a potential packing of E2 domains ( clusters 8 , 10; Fig 3C and 3D ) . Thus , BIS proposed that interdependent rearrangements of E1 and BL could represent a hallmark of E1E2 fusogenic conformational changes . Analysis of clusters from another HCV genotype ( gt2 ) provided similar findings as fusion clusters also involved the BL ( cluster 10 , 13 ) as well as distant blocks on E1 ( cluster 6 , 10 ) and E2 ( cluster 10 , 16 ) ( Fig 3E; S6–S8 Figs; S7–S9 Tables ) . Statistically significant coevolution networks between E1 and BL were also found among genotype 3 sequences ( S9 Fig ) . Similar networks were also found among sequences of genotype 4 to 6 , but displayed poor statistical significance for the reason described above . Genotype 3 to 6a cluster positions are available through the webpages indicated in the data availability statement . HCV E1 and E2 transmembrane were previously shown to be critical for E1E2 heterodimerization and correct E1E2 functions [26–28] . Consistently , BIS was also able to identify several coevolution clusters between the transmembrane of E1 and E2 using sequences of genotype 1 and 2 ( S10 Table ) , hence strengthening the functional significance of BIS analysis . In parallel , BIS also revealed several gt1a and gt2 structural and multifunctional clusters as supportive of the E2core central scaffold structure ( S5 and S8 Figs ) , reinforcing BIS as a relevant method to model viral protein conformations . The detailed analysis of all the clusters of gt1a and 2 , regardless of their attributed function , can be found in S1 and S2 Texts respectively . Altogether , BIS coevolution analysis of E1E2 sequences suggested that E2 may adopt a pre-fusion structure distinct from E2core as well as yet unreported molecular rearrangements that could occur during fusion . We hence hypothesized that a movement of the BL ( green; Fig 4A ) , through dialogs with E1 , could mediate the evolution from a potential stretched E2 pre-fusion structure toward a domain-packed E2 post-fusion structure ( Fig 4B ) . To challenge the potential role of the BL in E1E2 rearrangements , we generated a soluble 9kDa 6His-tagged BL domain ( 71 aa; named BLd-H77; Fig 4C; S10A Fig ) from the H77 gt1a strain , detectable through Coomassie blue staining and Western immunoblotting ( Fig 4D and 4E ) . Non-reducing SDS-Page electrophoresis and Dynamic Light Scattering ( DLS ) analysis ( S10B and S10C Fig ) confirmed the homogeneity of the peptide in solution and suggested that BLd-H77 fold as a monomer . The far UV circular dichroism ( CD ) spectrum of BLd-H77 eluted from size exclusion chromatography displays the molar ellipticity per residue expected for a protein folded mainly in α-helix ( S10D Fig ) . We next assessed its effect on HCV infection . Interestingly , BLd-H77 was able to inhibit , in a dose-dependent manner , infection of Huh7 . 5 cells by replicative hepatoma cell line-derived HCV particles ( HCVcc ) harboring envelope glycoproteins of gt1a ( H77/JFH-1 ) but also gt2a ( JC1 ) ( Fig 4F; S11A Fig ) . Although BLd-H77 was able to slightly inhibit HCVcc infection when pre-incubated with cells prior infection , it showed a strong potency to inhibit infection when present during the first four hours of infection ( Fig 4G ) thus suggesting that BLd-H77 might likely act on early steps of the virus life cycle . Consistently , BLd-H77 was able to efficiently inhibit infection of Huh7 . 5 by non-replicative retroviral pseudoparticles harboring HCV E1E2 ( HCVpp ) from different genotypes ( Fig 4H and 4I; S11B Fig ) . This inhibition was specific as BLd-H77 was not able to inhibit infection by pseudoparticles harboring VSVG envelope ( VSVGpp ) . Time-course experiments using BLd-H77 ( Fig 4I ) as well as two additional entry inhibitors , Bafilomycin A1 that acts on cell endosome acidification [29] ) and an anti-E2 neutralizing antibody ( that binds to E1E2 complexes [30] ) , demonstrated that BLd-H77 is an entry inhibitor that likely acts on viral particles ( S11C Fig ) but not on cells . No effect of BLd-H77 on HCV cell entry receptors expression could be observed ( S11D Fig ) . Moreover , BLd-H77 was also able to inhibit cell-to-cell transmission ( S11E Fig ) in addition to cell-free infection ( Fig 4F and 4G ) . Altogether , these results highlighted that BLd-H77 likely inhibits a conserved mechanism during HCV entry without affecting cell susceptibility for infection . Importantly , BLd-H77 was able to inhibit HCVcc and primary human hepatocytes-derived HCVcc virus ( or HCVpc ) infection of primary human hepatocytes ( PHH ) and Huh7 . 5 respectively ( Fig 4J and 4K; S11F Fig ) . Finally , we assessed the ability of BLd-H77 to inhibit infection in vivo . BLd-H77 showed a potency to inhibit HCVcc JC1 infection over time in humanized liver mice treated with 150 μg of BLd-H77 under a prophylactic protocol ( Fig 4L ) . Despite their uneven infectivity [31] , our results also suggested that BLd-H77 is able to inhibit patient-derived HCV particles infection in humanized mice p = 0 . 02 for all quantifiable values equal and above the detection limit ) as well as in PHH ( S11G and S11H Fig ) . BLd-H77 had no impact on human hepatocyte viability in mice as assessed by serum albumin concentration over the course of infection ( S11I Fig ) . Altogether , our results indicate that BLd-H77 is able to inhibit entry of different types of HCV particles in vitro and in vivo and thus target a strongly conserved virus entry mechanism . We then sought to elucidate how BLd-H77 blocks HCV entry . By pre-incubating viral particles with BLd-H77 and diluting the mix prior to infection to reach a BLd-H77 concentration below its efficient neutralizing activity ( determined in Fig 4F and 4H ) , we showed that BLd-H77 could irreversibly neutralize HCV particles regardless of mix dilution , hence suggesting that BLd-H77 can bind native viral particles prior viral entry ( Fig 5A ) . To assess the presence of an interaction between HCV particles and the BLd-H77 , we constructed a transmembrane form of BLd-H77 ( called BLd-tm ) ( Fig 5B ) . Following lentiviral transduction , BLd-tm expression was detectable at Huh7 . 5-BLd-tm surface ( S12A Fig ) , and did not impact HCV receptor expression ( S12B Fig ) . BLd-tm expression at Huh7 . 5 surface , but not expression of a similar construct encoding for an anchored HIV-1 fusion inhibitor ( namely C46 ) , inhibited HCVcc propagation both in a cell-free and cell-to-cell transmission manner ( Fig 5C and 5D; S12C Fig ) . HCVpp entry , but not VSVpp entry , ( Fig 5E ) was inhibited following infection of Huh7 . 5-BLd-tm , hence highlighting that BLd-H77 specifically inhibits HCV entry likely through binding of E1E2 glycoproteins . Consistently , more HCVpp were detected at Huh7 . 5-BLd-tm cell surface 4h post infection in comparison to Huh7 . 5 ( S12D Fig ) , suggesting a potential containment of HCV particles by BLd-tm at the cell surface . The ability of recombinant soluble E2 ( sE2 ) to bind more efficiently Huh7 . 5-BLd-tm cells than Huh7 . 5 cells in a dose dependent manner ( Fig 5F ) further suggested that virus entry is inhibited through an interaction between E2 and BLd-tm . In order to explore more precisely a putative interaction between E2 and BLd-H77 , we designed an ELISA assay to quantify the ability of sE2 to be captured by coated BLd-H77 . Our result showed a significant ability of sE2 to bind coated-BLd-H77 and coated-anti-E2 antibody AR3B [32] but not a coated-mouse IgG isotype ( Fig 5G; S12E Fig ) . Altogether , these results strong suggest that BLd-H77 inhibit HCV entry by binding to E2 glycoprotein . Next , we explored which step of HCV entry is inhibited by BLd-H77 . BLd-H77 had no significant effect on attachment of HCVcc ( HCVcc JC1 , Fig 5H ) , HCVpp-H77 particles or soluble E2 ( S13A Fig ) on Huh7 . 5 cells . Using a highly specific and previously established HCVpp binding assay [21] , we confirmed that BLd-H77 does not abrogate HCVpp binding to either human CD81 or SR-BI when used at a highly neutralizing concentration ( S13B–S13D Fig ) . Moreover , BLd-H77 neutralizing activity was not competing with the activity of a neutralizing anti-E2 antibody known to inhibit viral particle binding [30] , and their use in combination showed a synergistic neutralization effect ( S13E Fig ) . Consistently , BLd-H77 could bind viral particles following their binding at the cell surface , and was shown to have its more potent neutralizing activity during post-binding steps ( Fig 5I ) . Using a cell-cell fusion assay , we showed that BLd-H77 could strongly inhibit cell-to-cell fusion in comparison to control envelope glycoproteins ( Fig 5J; S14A Fig ) . Importantly , cell-to-cell fusion was only inhibited when cells were incubated with BLd-H77 before low pH exposure that activate membrane fusion ( S14B Fig ) , underlining BLd-H77 ability to specifically binds E1E2 pre-fusion conformations . Finally , using a HCVpp fusion assay with liposomes , which are devoid of any receptors or cell factors , the BLd-H77 inhibited fusion in a dose-dependent manner ( Fig 5K , S14C Fig ) . Altogether , these results suggest that BLd-H77 specifically blocks E1E2 fusogenic rearrangements and the formation of post-fusion structures through binding to E2 protein , in accordance with BIS predictions . Beside highlighting potential E1E2 rearrangements during fusion , BIS can identify pairs of residues that need to mutate in concert to guarantee structural compensations and proper viral fitness . Indeed , we have previously demonstrated how HCV entry depends on strain-specific dialogs between particular E1 and E2 domains [21] . Thus , we aimed at addressing whether BIS is able to unveil specific dialogs between residues of E1 and the E2 BL that modulate HCV fusion . The BIS predictions identified the multifunctional gt2 cluster 5 ( S7–S9 Tables ) as an interesting candidate for supporting E1 and E2 dialogs , BL movements and transition from E1E2 pre-fusion to post-fusion states . This cluster , similar to gt1a fusion cluster 5 , linked two central blocks in E1 ( residues 104 , 105 , and 109 ) and one block in E2 BL ( residues 427 to 436; orange; Fig 6A ) . In order to challenge its potential role during fusion , we used cluster 5 blocks to guide the rational design of E1E2 chimeric constructs . The E1 region containing two cluster 5 blocks ( Fig 6B; Region 1 ) and E2 BL regions containing the other cluster 5 block ( Fig 6B; Region 2 ) or not ( Fig 6B; Region 3 , non-coevolving cluster 5 block as a control of specificity ) were swapped between two E1E2 heterodimers from different gt2 strains , one allowing efficient HCVpp entry ( J6 ) and another one mediating sub-optimal HCVpp entry ( 2b1 ) ( S15A and S15B Fig ) . All chimeras were similarly expressed and incorporated ( S15C Fig ) . Although J6 chimera carrying 2b1 cluster 5 regions ( J6-1/2 ) displayed a poor entry efficiency , similar to 2b1 parental glycoproteins , a 2b1 chimera carrying J6 cluster 5 regions ( 2b1-1/2 ) exhibited >10-fold improved entry ability ( Fig 6C ) . In contrast , 2b1 chimera carrying J6 Region 1 and 3 ( 2b1-1/3; Fig 6C ) were not optimal for entry , hence suggesting that the E1 cluster 5 blocks and the N-terminal half of the BL domain are involved in a dialog regulating virus entry . Production and titration of HCVcc particles harboring these different chimeric envelopes confirmed these results ( Fig 6D , S15D Fig ) . Interestingly , 2b1-1/2 also displayed improved ability to mediate cell-to-cell fusion ( Fig 6E ) as well as higher fusion efficiency at neutral pH than at acidic pH ( Fig 6F ) , suggesting that this chimera exhibited an E1E2 conformation already primed for fusion at neutral pH . In contrast , J6-1/2 chimera did not increase J6 fusion efficiency and abrogated E1E2 sensitivity to low pH at levels similar to those of 2b1 ( Fig 6E and 6F ) . Altogether , consistently with BIS predictions , these results suggest that conserved interplays between the central region of E1 and the N-amino-terminus region of the BL likely govern E1E2 fusogenic conformational states . Our results also support the maintenance of such dialogs through coevolution as they appeared to be mediated by genotype-specific regions of E1 and the BL . To extend our transfer between the bioinformatics identification of coevolving amino acid clusters to the functional linkage of these domains , we tested the ability of BIS to pinpoint specific amino acids located in other regions than BL and furthermore , in the context of another genotype than genotype 2 ( studied above ) . For this purpose , we sought to challenge a gt1a fusion cluster identified by BIS ( Fusion cluster 5 , S4–S6 Tables ) , which involved four residues within E1 central region ( position 78 and 113–115 ) and a domain of 10 amino acids ( 452–461 ) within the Stem region . We employed two poorly divergent genotype 1a E1E2 sequences , H77 and A40 , that displayed two E1 ( SI/GM; position 112 and 117 ) and one E2 ( D/N; position 462 ) amino acid differences located at the borders of the gt1a fusion cluster 5 blocks ( Fig 7A; S16A Fig ) . Unlike J6 and 2b1 , the level of functionality of H77 and A40 were relatively close ( 2 . 4x104 and 1 . 6x104 GFP i . u . per ml respectively ) despite being significantly different ( Fig 7B ) , hence making it challenging to predict the influence of residue swaps on envelope functionality . H77 chimera harboring both swapped E1 and E2 A40 residues significantly impacted HCVpp infectivity , although swapped E1 or E2 residues alone did not impact H77 functionality ( Fig 7B ) despite similar E1E2 expression and incorporation ( Fig 7C ) . Importantly , H77 chimeras harboring only the E1 or E2 A40 residues showed defect for cell-cell fusion compared to H77 , although fusion ability of the H77 chimera harboring both the E1 and E2 mutations were enhanced ( Fig 7D ) . Altogether , consistently with BIS predictions , our results suggest that these E1 and E2 residues −and to larger extend the E1 central region and the E2 stem− are part of a coevolving network that regulates the fusogenic properties of gt1a viral envelope .
Flaviviridae are cause of many health concerns worldwide . A better understanding of the molecular processes regulating the life cycle of these viruses is critical for the design of potent anti-viral therapies . By taking advantage of the high-mutation rate of these viruses , coevolution analysis represents a valuable approach to decrypt viral protein functional rearrangements and provide basis for their inhibition . Here , we employed a recent coevolution analysis method , BIS [22–24] to provide a proof of concept of such approach . Coevolution signals detected within DENV E and Pr recapitulated several structural features of DENV E/E-Pr protein complexes in different conformational states , hence highlighting the structural accuracy of BIS predictions . Coevolution analysis of HCV E1E2 sequences from several genotypes and sub-types led to the identification of several coevolution signals in HCV E1E2 and suggested that E1 and E2 are strong coevolving partners that refold interdependently during fusion ( Fig 8 ) . Importantly , the E2 BL emerged as a key element of these rearrangements that could mediate the transition of E1E2 complex from a pre-fusion to a post-fusion conformation . We propose that during this transition , the endogenous E2 BL packs with the front sheet of E2 . Thus , a recombinant soluble BLd-H77 could compete with the endogenous E2 BL and block HCV fusogenic rearrangements , then leading to inhibition of membrane fusion ( Fig 8 ) . Such competition is consistent with the idea that E2 may harbor a stretched pre-fusion structure exposing internal epitopes . The existence of other E2 structures is also supported by the ability of neutralizing antibodies to target an epitope that is not exposed in the E2core structure ( i . e . aa305-324 [33] ) . Whether the entire BLd-H77 or specific BL amino acids regions are sufficient to inhibit HCV fusion remain to be determined . However , a recent study aiming to screen E2-derived peptide inhibiting HCV infection did not identify peptides within the BL [34] , suggesting that a large fraction of the BL , instead of a specific amino acid region , is likely to be required to inhibit HCV entry . This hypothesis reinforces the idea of a physical , but not functional , competition between the endogenous BL and BLd-H77 . Our work also suggests that the E2 BL acts in close collaboration with E1 and that these domains are probably in close proximity in E1E2 heterodimer . Indeed , E2core structure is highly concealed by glycans at the exception of the BL [7] . This , combined with the fact that E1 and BL do not seem to represent preferential targets of anti-HCV neutralizing antibodies , suggests that the BL region and E1 may conceal respective epitopes . The critical interplay between BL and E1 during fusion was also demonstrated at the amino acid level , as BIS was able to accurately identify coevolution signals between specific residues of E1 and the BL that tightly regulate HCV fusion ( Fig 8 ) . By employing an original multi-disciplinary approach combining computational analysis and experimental assays , our work sheds light on potentially important structural and functional features of the HCV fusion mechanism . Although the clear roles of E1 and E2 during the HCV fusion process still remain to be better defined through the structural resolution of the E1E2 heterodimer at different pH , our work initiates a path toward an experimentally-supported model for HCV-cell fusion . In this putative model , E1 would play the role of a fusion protein protruding onto the virus surface whereas E2 would be a receptor binding protein and a fusion chaperone concealing E1 epitopes . Although E2core is a truncated E2 protein and may not exhibit full E2 properties , E2core does not respond to pH variation [6] . This , combined with the fact that E2 does not harbor fusion protein structural features , could suggest that E2 needs to be associated with E1 to undergo conformational changes and to chaperone E1 fusion-promoting rearrangements . Interestingly , non-conserved E1 residues swapped between the J6 and 2b1 envelope ( Fig 6 ) were located at a very close proximity upstream of the putative E1 fusion peptide ( CSALYVGDLC ) [11–16] . This result may suggest that the interaction of these E1 residues with E2 BL could participate to critical E1-E2 rearrangements leading to the fusion peptide insertion into the endosomal membrane . Following receptor priming [35] and insertion of a putative E1 fusion peptide into the endosomal membrane , HCV fusion could be triggered by a fold-over of the C-terminal domain of E1 toward its N-terminal domain . This rearrangement could be mediated by E2 BL movements , or alternatively , could promote E2 BL movements . Overall , this interdependent refolding would result in the packing of the BL and of the E2 C-terminal region toward the E2 front sheet , and to the fusion of the host and viral membranes ( Fig 8 ) . The Ig-fold β-sandwich structure of E2core has been proposed to display similarities with domain III or B from class II fusion proteins [7] . By chaperoning E1 fold-over and membrane fusion , our results support that E2 function could be related to a domain III-like , as recently suggested [36] . Further structural studies aiming to resolve the full structure of the entire HCV heterodimer in pre- and post-fusion conformations will be needed to decipher the integral fusogenic rearrangements . Despite the fact that the structural differences detected for pestivirus E2 vs . HCV E2 glycoproteins makes unlikely that these viruses harbor a similar fusion mechanism , it is possible that their E1 proteins could be derived from a common ancestor and represent so far a potential new class of fusion protein as suggested by others [9 , 18] . Thus , although the resolution of the E2core structure has undermined the role of E2 as a fusion protein , HCV fusion is likely mediated by two interdependent partners that display original structures and conformational changes . To our knowledge , our data provide unique evidence that a component ( in this case , the BL ) derived from a receptor-binding protein with no fusion peptide can modulate viral fusion rearrangements , hence highlighting HCV fusion as a unique mechanism among known enveloped viruses . A primary limitation of current coevolution analysis approaches relies on the availability of a large number of sufficiently divergent evolutionarily-related sequences [37] . Such sets of sequences constitute the bottleneck for today’s coevolution analysis methods . We have shown previously that BIS can overcome these limitations and can address coevolution of conserved sequences such as viral genotype sequences [22–24] . Consistently , we failed to detect coevolution signals within our sets of E1E2 sequences when employing other existing coevolution methods ( such as DCA , PSICOV and EVcouplings [38–40] ) , methods that we previously discussed in [23] . In strong contrast to our previous work [23] that was critical to understand the computational power of coevolution analysis applied to viral sequences , our current study goes beyond the speculative and computational predictions and is fundamental is several ways . First , exploring coevolution signals when no or little structural and functional information are available remain highly challenging and hamper the delineation of undescribed viral protein-mediated processes . As coevolution signals imply structural or functional associations between coevolving residues [22 , 23] , our work shows that the BIS method is able to successfully highlight E1E2 contacts that likely orchestrate structural rearrangements of the heterodimer complex . In this respect , our work constitutes an important demonstration of BIS ability to decode structural features of major conformational change in proteins families characterized by few and conserved sequences . Hence , our work provides evidence that computational analysis of coevolution with BIS can be fruitfully used to find direct ( and possibly indirect ) contacts between proteins where the three-dimensional structure and rearrangements of protein complexes are not known . Second , it demonstrates that coevolution analysis can highlight the existence of conformational changes in proteins through pairs of coevolving residues that are not in contact in the known crystal structure . Many of the coevolution analysis tools developed in recent years ( such as DCA and DCA-like approaches ) are detecting “direct contacts” and justify their success by using the 3D crystal as their evidence of true positive predictions . In our study , we show that this idea only represents a part of the truth . Proteins are more complex systems , undergoing different structural conformations during their lifetime , and that evolutionary signals code not only for direct interaction but also for intermediate folding states and alternative structural conformations . Third , we advance in the comprehension of how computational techniques can be used to help revealing protein “contacts” that are biologically interesting , within and among structures . Finally , we provide experimental evidence of the biological significance of the coevolution signals for viral genotype sequences , hence allowing for a rapid identification of critical residue contacts regulating protein functions and conformations . Beyond uncovering important features of the HCV fusion mechanism , our study provides altogether a proof of concept that coevolution can be successfully harnessed to decrypt viral protein rearrangements and interactions , as well as to expand our knowledge of viral proteins-mediated biological processes . Virus entry mechanisms and envelope glycoproteins conserved epitopes represent valuable targets for the development of drugs and innovative vaccine strategies against challenging human pathogens [41–46] . By unscrambling key protein interactions or rearrangements , our work demonstrates that coevolution predictions can be of considerable value for stimulating a fast-tracked design and screening of innovative translational approaches and antiviral strategies unimpeded by virus plasticity .
Experiments were performed in accordance with the EU guidelines ( Directive 2010/63/EU ) on approval of the protocols by the local ethical committee ( Authorization Agreement C2EA-15; Ethic commitee: Comité d’Evaluation Commun au Centre Léon Bérard , à l’Animalerie de transit de l’ENS , au PBES et au laboratoire P4 ( CECCAPP ) , Lyon , France . Human Huh-7 . 5 ( kind gift of C . Rice , Rockefeller University , NY ) , BRL3A rat hepatoma ( ATCC CRL-1442 ) , and 293T kidney ( ATCC CRL-1573 ) cells were grown in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% fetal calf serum ( Invitrogen ) . Primary human hepatocytes ( PHH; BD Biosciences ) were centrifuged using a F-12 HAM medium ( Sigma ) and seeded overnight in collagen-coated 48 well plate ( 1x105 cells/well ) into Gentest seeding medium ( BD Biosciences ) completed with 5% FCS . The next morning , PHH were washed and cultured with the culture medium for PHH William’s E medium ( W4128 , Sigma ) supplemented with 7 , 5% BSA , 1% ITS ( insulin transferrin selenium , Gibco ) , 10-7M of Dexamethasone ( Sigma ) , 1% of non-essential amino acids ( Gibco ) , 1% of Glutamax ( Gibco ) and 1% Penicillin-Streptomycin solution ( Gibco ) . Sera containing HCV particles were obtained from a gt1b infected patient ( Hôpitaux Universitaires de Strasbourg , Strasbourg , France ) and were subsequently amplified in uPA-SCID humanized liver mice . Viral loads ( RNA copies number/ml ) were determined by RT-qPCR using a clinical diagnostic kit ( Abbot ) . The rat anti-E2 clone 3/11 [47] , the mouse anti-E2 AP33 ( Clayton et al . , 2002 ) and the conformational mouse anti-HCV E2 H53 [48] are kind gifts from J . McKeating ( University of Birmingham , UK ) , Arvin Patel ( MRC—University of Glasgow Centre for Virus Research , Glasgow , UK ) and J . Dubuisson ( Institut Pasteur de Lille , FR ) respectively . AR3B [32] and AR4A [30] antibody are a kind gift from Mansun Law ( The Script research institute , San Diego , USA ) . MLV Capsid was detected by a goat anti-MLV-CA antibody anti-p30 ( Viromed ) . HCVcc foci forming units were stained with a mouse anti-HCV NS5A antibody 9E10 [49] , a kind gift of C . Rice ( Rockefeller University , NY , USA ) . Human CD81 were detected with JS81 mAb ( BD Biosciences ) , human SRB1 with CLA-1 mAb ( BD Pharmingen ) , human Claudin-1 with the MAB4618 mAb ( R&D Syst . ) and human Occludin with an anti-Occludin mAb targeting the C-terminal region of the protein ( Laboratories Inc . ) . A Mouse anti-Human IgG2 antibody targeting human IgG2 hinge region ( Novus Biological ) was used to quantify BLd-tm expression at cell surface . A rabbit 6His-tag antibody ( Pierce antibody ) was used to detect soluble E2 and BLd-H77 6His-tag . In order to construct a relevant BLd soluble peptide derived from a genotype 1a sequence , we took into account the BL borders previously suggested ( Fig 3B , see also reference [7] ) , but also aimed to refine these borders using the locations of BIS coevolution networks of genotype 1a . Indeed , BIS suggested that a 15 amino acids extension from residue 390 to 405 , classified previously as an undefined domain [7] between the central Ig scaffold and the E2 BL , contained blocks ( S4 Fig , block 7–1 ) that coevolve with another region of the BL ( block 7–2 ) , forming altogether a fusion cluster ( cluster 7 ) . Hence , in order to rigorously challenge BIS predictions that suggested a potential redefinition of the BL , we constructed a soluble BLd peptide from residue 390 to 460 . Hence , DNA sequence of HCV E2 coding for the residues 390 to 640 was amplified from a genotype 1a envelope H77 ( AF009606 ) cDNA sequence . This sequence was subcloned into a phCMV plasmid to fuse the last 18 amino acids from the C-term part of HCV core that act as a signal peptide , and a CH3-terminal 6 Histidine-tag was added . The resulting peptide of 77 amino acids ( 71+6 ) was then named as BLd-H77 . BLd-H77 was expressed in 293T following transient transfection and purified from cell culture supernatant ( OptiMEM ) by fast protein liquid chromatography on a Superdex G-75 gel filtration column ( GE Healthcare ) . BLd-H77 was re-suspended in 1XPBS . The concentration of purified BLd-H77 peptides was determined by absorption at 205 nm . The mass of BLd-H77 peptide was measured by ESI mass spectrometry using a Finnigan LCQ ion trap mass spectrometer ( Thermo Electron Corporation ) . Analysis by SDS-PAGE electrophoresis was performed using a standard Tris–glycine system and 11% acrylamide gels , in reducing or non-reducing condition . Electrophoresis was followed by either coomasie blue staining or western immunoblotting with an anti His tag antibody . An anchored form of BLd-H77 was engineered by adding a hinge region ( human IgG2 ) and the transmembrane domain of the CD34 protein to the E2 BLd . The construct , named BLd-tm , was then inserted into Gae-SFFV-IRES backbone harboring selectable marker gene P140K MGMT . A similar construct , but encoding for a HIV-1 gp41 fusion inhibitor peptide [50] , namely C46 , was used as a control . Construct details are available upon request . Lentiviral vectors transducing BLd-tm or C46 were produced from 293T cells . Stable expression in Huh-7 . 5 was obtained by transduction with vector particle-containing supernatants of 293T producer cells , followed by selection of O6-benzylguanine and BCNU . BLd-tm expression in Huh-7 cells was quantified by FACS analysis using a mouse anti human IgG2 and an anti-mouse APC antibody . BRL cells expressing CD81 or SRBI were washed and stained for 1h at 4°C with a mouse anti-human IgG2 antibody ( for detection of C46 and BLd-tm ) , an anti-CD81 ( JS81 ) and with an anti-human SR-BI ( CLA-1 ) respectively . Cells were then washed and incubated with a secondary anti-mouse or rat APC antibody for 1h at 4°C . Cell surface expression levels were then quantified by flow cytometry ( FACS CANTO II–BD Biosciences ) . For HCV receptor detection on Huh7 . 5 and Huh7 . 5-BLd-tm cells , cells were fixed with 2% formaldehyde for 20 min at room temperature and washed . For Occludin staining , cells were permeabilized with Perm/Wash Buffer ( BD Biosciences ) for 15 min at 4°C prior staining . Human CD81 was stained with JS81 mAb , human SRBI with CLA-1 mAb , human Claudin-1 with the MAB4618 and human Occludin with an anti-Occludin mAb targeting the human Occludin C-terminal region . To characterize the effect of BLd-H77 on HCV receptor expression , Huh7 . 5 were incubated overnight with BLd-H77 ( 50μg/ml ) prior staining . HCVpp were produced as previously described [21 , 51] from 293T cells cotransfected with a murine leukemia virus ( MLV ) Gag-Pol packaging construct , an MLV-based transfer vector encoding the green fluorescent protein , and E1E2 envelope expression constructs H77 ( AF009606 ) , A40 ( unreferenced ) , UKN1B 12 . 6 ( AY734975 ) , UKN2A 2 . 4 ( AY734979 ) , JFH-1 ( AB047639 ) , J6 ( AF177036 ) , UKN2B 2 . 8 ( AY734983 ) , UKN3A 1 . 9 ( AY734985 ) , UKN4 21 . 16 ( AY734987 ) , UKN5 14 . 4 ( AY785283 ) , HK 6A-2 . 1 ( FJ230883 ) or control envelope HA-NA ( CY077420 ) and VSV-G ( AJ318514 ) . All chimeric J6/2b1 E1E2 heterodimers were constructed by molecular cloning , PCR and/or digestion between the genotype 2a envelope J6 ( AF177036 ) and a genotype 2b envelope UKN-2b1 ( unreferenced ) . All chimeric H77/A40 E1E2 heterodimers were constructed using a similar strategy between the genotype 1a envelope H77 ( AF009606 ) and A40 ( unreferenced , but previously employed [21] ) . 72 to 96h following infection , percentage of infected cells was quantified by FACS Canto II or LSRII ( BD Biosciences ) to quantify GFP expression . For BLd-H77 dose-dependent neutralization assay , HCVpp H77 or VSVGpp were pre-incubated with different concentrations of BLd-H77 or with PBS for 1h at room temperature and were then used to infect Huh7 . 5 . For the time-course neutralization assay , Huh7 . 5 were infected with HCVpp H77 for 4h prior washing . PBS , BLd-H77 ( 50 μg/ml ) , Bafylomycin A1 ( 20nM ) or AR4A ( 25 μg/ml ) were added into cell supernatant for 1h prior infection , during infection or after infection . Percentage of infected cells was determined 72h following infection . For HCVpp co-neutralization assay , HCVpp H77 pseudoparticles were pre-incubated for 1h at room temperature with BLd-H77 alone ( 35 or 50 μg/ml ) , AR4A alone ( 2 or 17 μg/ml ) or with both BLd-H77 and AR4A ( 35 and 2 μg/ml , or 35 and 17 μg/ml ) . Pre-mixes were then used to infect Huh7 . 5 and media was changed 6h post infection . GFP intracellular levels were quantified 4 days post infection by flow cytometry . For HCVpp post-binding neutralization assay , HCVpp-H77 pseudoparticles were incubated with Huh7 . 5 in presence of BLd-H77 ( 50 μg/ml ) or AR4A ( 25μg/ml ) for 1h at 4°C ( binding ) , for 4h at 37°C following binding ( entry ) , or for 72h following entry ( post-entry ) . As control , Huh7 . 5 were incubated and infected with HCVpp-H77 in a similar manner but not treated with BLd-H77 or AR4A . GFP intracellular levels and related percentage of infection were then quantified by flow cytometry . For HCVpp containment assay on Huh7 . 5-BLd-tm , Huh7 . 5 or Huh7 . 5-BLd-tm were infected or not with HCVpp H77 for 5h . Then , cells were washed and E2 cell surface expression was determined by flow cytometry following staining using the anti-E2 H53 antibody and a secondary anti-mouse APC antibody . Huh7 . 5 and Huh7 . 5-BLd-tm were infected similarly and GFP expression of infected cells was analyzed 72h post-infection . Transfected 293T cells were lysed and nuclei were removed by centrifugation at 12 000 rpm for 10 min . HCVpps were purified and concentrated from the cell culture medium by ultracentrifugation at 82 , 000xg for 1h 45 min through a 20% sucrose cushion . Cell lysates and viral pellets were subjected to western blot analysis using 3/11 anti-E2 antibody and an anti-MLV-CA antibody as described previously [21] . Plasmid pFK H77/JFH1/HQL ( kind gift of R . Bartenschlager ) , termed as H77/JFH-1 , displaying HCV genome with adaptive mutations ( Y835H in NS2 , K1402Q in NS3 , and V2440L in NS5A ) and harboring H77 sequence derived from the BLd-H77peptide , as well as plasmid pFKi389-Venus-Jc1 , termed as Jc1 , ( an intra-genotypic recombinant between J6-CF sequence ( AF177036 ) and JFH1 sequence ) were used to produce and electroporate into Huh7 . 5 the respective H77/JFH-1 and JC1 viral RNAs as described previously [21] . Huh7 . 5 cells , Huh7 . 5-BLd-tm or Huh7 . 5-C46 were infected with different dilutions of culture supernatants harvested at 24h , 48h and 72h post electroporation . Four days post-infection , cells were fixed with EtOH 100% and foci forming units ( FFUs ) were visualized after NS5A immunostaining as described previously [21] . For BLd-H77 dose-dependent neutralization assay , HCVcc particles were pre-incubated with different concentrations of BLd-H77 or with PBS for 1h at room temperature and were then used to infect Huh7 . 5 . For time-course neutralization assay , Huh7 . 5 were infected with HCVcc for 4h prior washing . PBS or BLd-H77 ( 35 μg/ml ) were added into cell supernatant for 1h prior infection , during infection or after infection . FFU/ ml were determined 4 days post-infection . To construct HCVcc particles harboring 2b1 , J6-1/2 , 2b1-1/2 or 2b1-1/3 envelope , we inserted by molecular cloning the related envelope into the pFKi389-Venus-Jc1 molecular clone , that initially encodes for J6 envelope . Viral RNAs were electroporated into Huh7 . 5 as described above . At 72h post electroporation , cell culture supernatants were tittered and used to infect naïve Huh7 . 5 . Number of foci forming units per ml were determined 4 days post infection as described above . In parallel , viral RNA were extracted from electroporated cell culture supernatants at 72h post ( ZR viral RNA kit , Zymo ) . HCV viral RNA copy number was quantified by one-step reverse transcription-PCR ( RT-PCR ) using MultiCode-RTx Real-Time PCR ( Luminex ) according to manufacturer’instructions and run on a Step One Plus quantitative PCR machine ( Life Technologies ) . Data were analyzed using the MultiCode Analysis Software v1 . 6 . 5 ( Luminex ) . The following primers were used for the detection of HCV RNA: GCTCACGGACCTTTCA ( sense ) and GGCTCCATCTTAGCCC ( antisense ) . H77/JFH1 virus was used to infect Huh7 . 5 and Huh7 . 5-BLd-tm ( m . o . i . 0 , 1 ) . At day 1 , 3 and 5 post infection , cells were fixed with 2% formaldehyde for 20 min at room temperature , washed and permeabilized with Perm/Wash Buffer ( BD Biosciences ) for 15 min at 4°C . NS5A expression levels were then quantified using anti-NS5A antibody 9E10 by flow cytometry ( FACS CANTO II–BD Biosciences ) . In parallel , cell supernatants were harvested at each time point , filtered and used to infect naïve Huh7 . 5 . 72h post infections , infectious titers ( FFU/ml ) of cell supernatant were determined as described above . HCVpp H77 and H77/JFH-1 HCVcc particles were preincubated for 1h at room temperature with PBS or BLd-H77 ( 50 μg/ml or 35 μg/ml respectively ) . Then , BLd-H77 and concentrated viral particles were diluted ( 1/5 ) with cell culture media or not prior infection of Huh7 . 5 . 4 days post infection , percentages of GFP-positive cells were determined by flow cytometry and FFUs/ml were determined by NS5A immunostaining as described above . H77/JFH-1 were used to infect Huh7 . 5 for 4h at 37°C ( m . o . i . 0 , 05 ) . After washing , cells were incubated with anti-E2 antibody AP33 ( 25μg/ml ) alone , or mixed with BLd-H77 ( 35μg/ml ) , or with PBS . 72h post infection , cells were fixed and numbers of cell per foci for each condition were quantified through NS5A immunostaining as described above . PHH were washed and infected by JC1 HCVcc virus at different m . o . i . ( 0 , 005; 0 , 001; 0 , 05; 0 , 1 ) . 4 days post infection , cell culture supernatants were harvested and used to infect naïve Huh7 . 5 . Infectious titers were revealed through NS5A immunostaining 4 days post infection as described above . Indirect titrations were performed as infected PHH were poorly detectable through NS5A immunostaining , making the indirect titration the only accurate method to quantify the amount of infectious viral particles ( and not physical viral particles ) in a PHH-cell culture supernatant . For neutralization assay , JC1 virus or JC1-derived HCVpc were pre-incubated with BLd-H77 ( 10 , 20 and 40 μg/ml ) or with PBS for 1h prior PHH ( m . o . i . 0 , 05 ) or Huh7 . 5 infection respectively ( m . o . i . 0 , 01 and 0 , 02 ) . 4 days post infection , infectious titers of PHH cell culture supernatants were determined by infecting Huh7 . 5 as described above . In parallel , infectious titers of JC1-derived HCVpc following Huh7 . 5 infection were quantified as described above . Sera containing HCV particles were used to infect PHH at a m . o . i . of 0 , 1 following BLd-H77 ( 30 μg/ml ) or PBS preincubation for 1h at room temperature . 4 days post infection , cell culture supernatants were harvested and viral loads ( RNA copies number/ml ) were determined by RT-qPCR using a clinical diagnostic kit ( Abbot Real Time™ HCV assay ) with a limit of quantification ( LOQ ) of 12 IU/mL ( i . e . 51 . 6 HCV RNA copies/ml ) . Given a serum dilution of 1:100 in PBS , LOQ = 5160 HCV RNA copies/ml . Retroviral vectors expressing human CD81 ( NM_004356 ) and SR-BI ( Z22555 ) were described previously [52] . Retroviral vectors containing these cDNAs were produced from 293T cells as VSV-G pseudotyped particles as described previously [53 , 54] . Stable expression of either receptor in BRL cells was obtained as described previously [52] . Binding assays were performed as described previously [21] . Briefly , 50μl of concentrated virus ( 100x ) or 100 ul of concentrated soluble E2 ( 100x ) were pre-incubated with BLd-H77 ( 50 μg/ml ) or with PBS for 1h at room temperature . Then , pseudoparticles or soluble E2 were mixed with Huh7 . 5 , BRL , BRL-CD81 or BRL-human SR-BI in presence of 0 . 1% sodium azide for 1h at 37°C . Cells were then washed with PBFA ( PBS , 2% fetal bovine serum , and 0 . 1% sodium azide ) . Bound viruses were detected using the mouse H53 anti-HCV E2 antibody and soluble E2 was detected using either the H53 antibody or a rabbit 6His-tag antibody for 1h at 4°C . After washing , primary antibodies were quantified by flow cytometry ( FACS Canto II , BD Biosciences ) using APC goat anti-mouse immunoglobulin-G . In parallel , concentrated HCVpp were pre-incubated with 50 μg/ml of BLd-H77 and used to infect Huh7 . 5 in order to verify the neutralizing effect of BLd-H77 on the entry of concentrated HCVpp . For BLd-H77 binding assay , BLd-H77 ( 50 μg/ml ) or PBS were mixed with Huh7 . 5 for 1h at 37°C prior 6His-tag staining using a rabbit 6His-tag antibody . Concentrated soluble E2 ( 100x; 100 or 250 μl ) were mixed with equivalent number of Huh7 . 5 or Huh7 . 5-BLd-tm ( 2x105 or 5x105 cells ) in presence of 0 . 1% sodium azide for 1h at 37°C . After washing , bound soluble E2 were detected using anti-E2 H53 antibody as described above . Levels of binding enhancement were determined relatively to the basal E2 binding on naïve Huh7 . 5 . For HCVpp binding enhancement assay , 1x105 Huh7 . 5 cells or Huh7 . 5-BLd-tm were infected with HCVpp for 4 hours . Cells were then trypsinized and washed with PBFA ( PBS , 2% fetal bovine serum , and 0 . 1% sodium azide ) . Bound viruses were quantified by flow cytometry ( FACS Canto II , BD Biosciences ) following cell staining with the mouse H53 anti-HCV E2 and an APC goat anti-mouse immunoglobulin-G . JC1 HCVcc particles ( 5x104 i . u . ) were pre-incubated with BLd-H77 ( 35 μg/ml ) , Heparine ( 250 μg/ml ) or with PBS for one hour at 37°C . Viral particles were then mixed with 1x105 Huh7 . 5 cells for 2h at 4°C . After 3 washings , cells were lysed using the RLT Buffer ( QIAGEN ) complemented with β-mercaptoethanol . Total RNAs were then extracted using the RNeasy mini kit ( QIAGEN ) as recommended by the manufacturer . Extracted RNAs were reverse transcribed using the iScript cDNA synthesis kit ( Bio-Rad ) and HCV viral RNA ( 5-CTTCACGCAGAAAGCGCCTA and 5-CAAGCGCCCTATCAGGCAGT ) and human GAPDH ( 5- GAAGGTGAAGGTCGGAGTC and 5- GAAGATGGTGATGGGATTTC ) were then quantified by qPCR using the FastStart Universal SYBR Green Master kit ( Roche Applied Science ) on an Step One Plus quantitative PCR machine ( Life Technologies ) . Cell-associated viral RNA copies were normalized on human GAPDH expression for each sample . 96-well plates ( Corning ) were coated overnight with different amounts of a mouse IgG isotype ( 10ng and 100ng; Abcam ) , anti-E2 antibody AR3B [32] ( 10 and 100ng ) , and BLd-H77 ( 10 , 100 and 250ng ) . The next day , following a one-hour incubation step with a SuperBlock blocking buffer ( Thermo Scientific ) to prevent non-specific binding , each coating condition was incubated with 10ng of sE2 or not . Interactions were revealed using a rat anti-E2 antibody 3/11 [47] or a rat IgG isotype , then followed by an incubation with an anti-rat HRP antibody ( Biorad ) . Optical density signals at 450 nm were then assessed using a TriStar Multimode Microplate reader ( Berthold ) . FRG ( Fah–/–Rag2–/–Il2rg–/– ) mice ( mixed background: C57BL/6 and 129Sv ) were housed in our animal facility ( Plateau de Biologie Experimentale de la Souris , PBES , Lyon , France ) . Because of their lethal phenotype , mice are maintained on 8 mg/l of NTBC ( nitro-4-trifluoro-methylbenzoylcyclohexanedione ) in the drinking water . 48h prior the engraftment , adult ( 6–10 weeks old ) mice were injected intravenously with 2x109 p . f . u . of an adenoviral vector encoding the uPA transgene . 7x105 to 1x106 PHH ( BD Biosciences ) were injected intrasplenically as previously described [55] . Immediately after engraftment , the NTBC was progressively withdrawn as follow: 2 days at 10% of colony maintenance concentration , then 2 days at 5% , 2 days at 2 . 5% , then the NTBC was completely removed . During the phase without NTBC , mice were weighted every two days . After 2–6 weeks , mice with clinical symptoms ( lethargy , hunched posture ) or severe weight loss ( >15% ) were put again on NTBC for 3 days before second withdrawal ( cycling ) . Cycling was repeated until clinical symptoms resolved . In order to prevent the development of a murine hepatocellular carcinoma , highly reconstituted mice selected for infectious experiments were subjected to further NTBC treatment ( 3–4 weeks w/o NTBC and 3 days with 100% ) . Blood from transplanted mice and controls were collected every 2–4 weeks after engraftment by retro orbital puncture . Sera were sent to a diagnostic laboratory for quantification of human Albumin ( Cobas C501 analyzer , ROCHE ) . Highly reconstituted mice ( HSA >15mg/ml ) were infected with JC1 HCVcc particles inocula ( 105 i . u . ) or with patient-derived HCV particles ( 5x104 i . u . ) via intraperitoneal route . Mice sera were collected at day 7 and day 14 post-infection by retro-orbital bleeding . At day 21 , mice were sacrificed , sera were collected and levels of human albumin were determined ( Cobas C501 analyzer , ROCHE ) . JC1 infectious titers in mice sera were determined through infection of Huh7 . 5 with different dilutions of sera as described above . In parallel , for HCV particle containing sera , viral load was determined by RT-qPCR using a clinical diagnostic kit ( Abbot Real Time™ HCV assay ) with a limit of quantification ( LOQ ) of 12 IU/mL ( i . e 51 . 6 HCV RNA copies/ml ) . Given a serum dilution of 1:100 in PBS , LOQ = 5160 HCV RNA copies/ml . A first cohort of 11 mice was attributed for an in vivo inhibition assay of JC1 infection ( 9 mice + 2 negative controls; one non-infected and one non-engrafted ) . 7 mice were treated under a prophylactic protocol with PBS ( 4 mice ) or with 30 μg ( 3 mice ) of BLd-H77 . Mice were treated via intra-peritoneal route one day prior infection , and at day 1 , 7 and/or 14 post-infection infection . For all the mice , sera were harvested and viral titers were quantified at day 7 , 14 and during sacrifice 21 post infection . A second independent cohort of 7 mice was attributed for another in vivo inhibition assay of JC1 infection ( 5 mice + 2 negative control ) . Here , JC1 infection was challenged with a dose of 150μg of BLd-H77 ( 2 mice ) or with PBS ( 3 mice ) under a prophylactic protocol similar as described above . A cohort of 14 mice ( 12 mice + 2 negative control ) was attributed for an in vivo inhibition assay of serum-derived HCV particles infection , challenged with one dose of BLd-H77 ( 150 μg , 5 mice ) or with PBS ( 7 mice ) under a prophylactic protocol similarly to what has been described above . Serum-derived HCV particles viral load were determined as described above . Following sacrifice , the level of human Albumin was quantified for each untreated and treated mice in order to ensure that HCV infection inhibition was not due to a decrease of liver humanization caused by BLd-H77 . Cell-cell fusion assays were performed as described previously [13 , 19] . Briefly , HEK-293T cells ( 2 . 5x105 cells/well seeded in six-well tissue culture dishes 24 h before transfection ) were co-transfected using calcium phosphate reagent with a HCV ( H77 , J6 , 2b1 or J6/2b1 chimera ) or VSV-G envelope encoding-plasmids and with an HIV-1 LTR ( long terminal repeat ) luciferase reporter plasmid . After 12h , transfected HEK-293T cells were detached with versene ( 0 . 53 mM EDTA; Invitrogen ) and co-cultured ( 5x104 cells/well ) with Huh-7-Tat indicator cells ( 5 . 104cells/well ) in a 24-well plate . After 24 h , the cells were washed with serum free DMEM , incubated for 3 min in either pH 7 or pH 5 buffer ( 130 mM NaCl , 15 mM sodium citrate , 10 mM Mes and 5 mM Hepes ) and then washed three times with serum free DMEM . The luciferase activity was measured 72 h later using a luciferase assay kit according to the manufacturer’s instructions ( Promega ) . For fusion neutralization assay , coculture were washed , pre-incubated with BLd-H77 ( 50μg/ml ) or with PBS for 1h at 37°C prior a second washing and exposure to pH buffer . Alternatively , co-cultured cells pre-treated by BLd-H77 or PBS were incubated for an extension of 24h following pH shock by BLd-H77 ( 50μg/ml ) or PBS . HCVpp/liposome lipid mixing was performed as previously described [21 , 56] . R18-labeled liposomes were obtained by mixing octa-decyl rhodamine B chloride ( R18; Molecular Probes ) and lipids ( phosphatidylcholine and cholesterol; Aventi ) and mixed with 40 μl of concentrated HCV pseudoparticles ( non-enveloped HCVpp , HCVpp-H77 and HCVpp harboring the fusion defective envelope E140/E2H77 previously characterized [21] ) or retroviral particles pseudotyped with Influenza Hemagglutinin-Neuraminidase ( HANApp ) , all diluted in PBS ( pH 7 . 4 ) within a 37°C thermostable 96-well plate . After pH decrease to 5 ( acidification ) , dequenching of R18 due to lipid mixing between HCVpp and liposomes were recorded on a micro-plate fluorometer ( InfiniteM1000 Tecan Group Ltd ) for a period of 5 to 20min with an excitation wavelength ( λexc ) at 560 nm and an emission wavelength ( λem ) at 590 nm . Maximal R18 quenching was measured after the disruption of liposomes by the addition of 0 . 1% TritonX-100 . For fusion-neutralization assays , pseudoparticles were pre-incubated with different doses of BLd-H77 or with PBS for 1h prior incubation with liposomes and acidification . Structural analysis of Dengue E pre-fusion structure ( PDB 1K4R ) , E-Pr complex structure ( 3C6E ) , E post-fusion structure ( 1OK8 ) and E2core structure ( 4MWF ) were realized using Chimera software [57] ( UCSF ) . GraphPad Prism software ( version 6 ) was used for statistical analysis . Statistics were calculated using Student’ t test and/or two-way ANOVA when appropriate . ( *p<0 . 05 , **p<0 . 01 , ***p<0 . 001 ) . Protein amino acid sequence alignments were realized using ClusterX2 . 1 and rendered as Post-script format . The BIS methodology and related-coevolution signal analysis are described in details in reference [22] , [23] and [24] . Applicability of BIS for detecting coevolution signals in viral sequences is specifically described in reference [23] . Literature describing amino acids mutations within E1E2 sequences impacting E1E2 folding/heterodimerization , E1E2 binding to cellular receptors or E1E2 fusion was used as references to attribute functions to clusters mapping specific regions within E1E2 sequences . The references used were the following: Folding/Heterodimerization [19 , 27 , 28 , 58–60] , Viral binding site conformation [19 , 20 , 26 , 60–66] and Fusion [11 , 13 , 19–21 , 27 , 67] . | Several virus-mediated molecular processes remain poorly described , which dampen the development of potent anti-viral therapies . Hence , new experimental strategies need to be undertaken to improve and accelerate our understanding of these processes . Here , as a proof of concept , we employ amino-acid coevolution as a tool to gain insights into the structural rearrangements of Hepatitis C Virus ( HCV ) envelope glycoproteins E1 and E2 during virus fusion with the cell membrane , and provide a basis for the inhibition of this process . Our coevolution analysis predicted that a specific domain of E2 , the Back Layer ( BL ) is involved into significant conformational changes with E1 during the fusion of the HCV membrane with the cellular membrane . Consistently , a recombinant , soluble form of the BL was able to inhibit E1E2 fusogenic rearrangements and HCV infection . Moreover , predicted coevolution networks involving E1 and BL residues , as well as E1 and BL-adjacent residues , were found to modulate virus fusion . Our data shows that coevolution analysis is a powerful and underused approach that can provide significant insights into the functions and structural rearrangements of viral proteins . Importantly , this approach can also provide structural and molecular basis for the design of effective anti-viral drugs , and opens new perspectives to rapidly identify effective antiviral strategies against emerging and re-emerging viral pathogens . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"flow",
"cytometry",
"cell",
"physiology",
"enzyme-linked",
"immunoassays",
"microbiology",
"viral",
"structure",
"membrane",
"fusion",
"coevolution",
"immunologic",
"techniques",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"sequence",
"alignment",
"bioinformatics",
"immunoassays",
"cell",
"membranes",
"spectrophotometry",
"cytophotometry",
"cell",
"biology",
"virology",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"evolutionary",
"biology",
"evolutionary",
"processes",
"spectrum",
"analysis",
"techniques",
"cell",
"fusion"
] | 2018 | A protein coevolution method uncovers critical features of the Hepatitis C Virus fusion mechanism |
Across the mammalian nervous system , neurotrophins control synaptic plasticity , neuromodulation , and neuronal growth . The neurotrophin Brain-Derived Neurotrophic Factor ( BDNF ) is known to promote structural and functional synaptic plasticity in the hippocampus , the cerebral cortex , and many other brain areas . In recent years , a wealth of data has been accumulated revealing the paramount importance of BDNF for neuronal function . BDNF signaling gives rise to multiple complex signaling pathways that mediate neuronal survival and differentiation during development , and formation of new memories . These different roles of BDNF for neuronal function have essential consequences if BDNF signaling in the brain is reduced . Thus , BDNF knock-out mice or mice that are deficient in BDNF receptor signaling via TrkB and p75 receptors show deficits in neuronal development , synaptic plasticity , and memory formation . Accordingly , BDNF signaling dysfunctions are associated with many neurological and neurodegenerative conditions including Alzheimer’s and Huntington’s disease . However , despite the widespread implications of BDNF-dependent signaling in synaptic plasticity in healthy and pathological conditions , the interplay of the involved different biochemical pathways at the synaptic level remained mostly unknown . In this paper , we investigated the role of BDNF/TrkB signaling in spike-timing dependent plasticity ( STDP ) in rodent hippocampus CA1 pyramidal cells , by implementing the first subcellular model of BDNF regulated , spike timing-dependent long-term potentiation ( t-LTP ) . The model is based on previously published experimental findings on STDP and accounts for the observed magnitude , time course , stimulation pattern and BDNF-dependence of t-LTP . It allows interpreting the main experimental findings concerning specific biomolecular processes , and it can be expanded to take into account more detailed biochemical reactions . The results point out a few predictions on how to enhance LTP induction in such a way to rescue or improve cognitive functions under pathological conditions .
Brain-Derived Neurotrophic Factor ( BDNF ) is a member of the protein family of mammalian neurotrophins , further comprising nerve growth factor , neurotrophin 3 and neurotrophin 4/5 . Neurotrophins are well known across the animal kingdom to support survival , ontogenetic development , differentiation , and stability of neurons in the entire nervous system[1 , 2] . In the mature nervous system , BDNF , in particular , serves additional roles by regulating functional and structural synaptic plasticity ( reviewed e . g . in [1 , 3–5] ) . In recent years , a wealth of data has been accumulated on the many roles of BDNF in regulating synaptic plasticity at glutamatergic and GABAergic synapses , e . g . in hippocampus , neocortex , amygdala , and cerebellum , unraveling signaling pathways of unprecedented complexity [1 , 6–10] . However , despite this well-established role of BDNF as a central activity-dependent mediator ( i . e . switching on biochemical pathways that induce and maintain enhanced synaptic transmission [11–13] ) and modulator ( i . e . , facilitating synaptic changes that are mediated by other signaling pathways [3 , 14–17] ) of synaptic plasticity , the interplay between intra- and extracellular signaling pathways [18 , 19] that regulate and fine-tune BDNF-dependent synaptic changes is not well understood . The overall picture is rather complex . BDNF consists of a protein homodimer that is generated exclusively in glutamatergic neurons from two identical peptide chains held together by noncovalent interactions . The precursor protein , pre-proBDNF , is sequestered into the endoplasmic reticulum , where the pre-sequence is cleaved off , yielding proBDNF . Intracellularly , proBDNF can be cleaved ( by protein convertases , PCs and furin ) into mature BDNF ( mBDNF ) and BDNF pro-peptide . All three BDNF species are thought to be assembled into secretory vesicles that are transported to the plasma membrane in soma , dendrites , and axons , where they release their content via Ca2+-dependent exocytosis [20] . Following secretion , remaining proBDNF can be cleaved by extracellular proteases ( e . g . plasmin and matrix metalloproteinases ) . This is an important functional step since at this point it is determined whether mBDNF or proBDNF dependent signaling cascades are activated at a synapse . Because proBDNF and mBDNF activate signaling cascades that partially antagonize each other , the importance of knowing the exact identity of released BDNF can hardly be overestimated . While mBDNF preferentially binds to the tyrosine-kinase receptor B ( TrkB ) and , among other functions , supports LTP , proBDNF preferably docks to the p75 receptor , which mediates long-term depression ( LTD ) [8] . The complexity of BDNF control over neuronal growth , plasticity , and modulation , makes it difficult to carry out experimental studies to fully understand BDNF-dependent processes . Computational modeling can significantly help to untangle the interplay of these processes but , despite the widespread implications of BDNF signaling in structural and functional neuromodulation during normal and pathological physiological conditions , a biologically realistic model of how BDNF signaling instructs these changes is still missing . Except for a very recent example of a model of a positive BDNF feedback loop , to take into account experiments on inhibitory avoidance training [21] , to the best of our knowledge there are no published models available that address BDNF-dependent pathways . In this paper , we set out to investigate BDNF-dependent synaptic mechanisms , by implementing the first kinetic model of the central BDNF-dependent subcellular pathways underlying spike timing-dependent Long-Term Potentiation ( t-LTP ) at hippocampal synapses . For this purpose , we focused on TrkB-dependent processes at hippocampal Schaffer collateral to CA1 pyramidal cell synapses , for which extensive experimental work is available that can be used to constrain the parameter values [11 , 13] . We show that the model can capture the main experimental findings by using a minimal set of subcellular pathways , with which we can make specific predictions on how to enhance LTP induction in such a way to rescue or improve cognitive functions under pathological conditions .
As a reference for our model , we considered the data from [11] . In the paper , the authors described t-LTP elicited in hippocampal pyramidal CA1 neurons by repeatedly pairing , with different delays ( Δt ) , a single stimulation of the Schaffer Collaterals with one , two or four postsynaptic action potentials elicited at 200 Hz . These induction protocols are hereof designated 1:1 , 1:2 , and 1:4 t-LTP . The paper highlights important properties of the mechanisms underlying t-LTP . Their main results are summarized in Fig 1A , where the EPSP slope recorded in whole-cell patch clamp mode is plotted as a function of time . On average , the expression of t-LTP was relatively delayed , and it took approximately 30 min to reach its maximum expression ( Fig 1A , data reproduced from Fig . 1B of [11] ) . The increase in synaptic strength after both 1:1 and 1:4 t-LTP was graded with time . Assuming that an individual synapse switches to a potentiated state following an all-or-none change [22 , 23] , this progressive increase in the overall t-LTP observed at the soma suggests a distribution of transition times for different spines , driven by the time course of the processes underlying t-LTP induction and expression . For both protocols , a Δt>15 ms did not result in a significant t-LTP ( Fig 1B ) , being consistent with other experimental findings [24 , 25] . As commonly expected , only short positive delays between pre- and postsynaptic stimulation are efficient to produce timing-dependent LTP , while longer delays reduce t-LTP magnitudes . In the experimental study that forms the basis of our model [11] a significant reduction of t-LTP was observed with positive time delays between 15–25 ms . It should be stressed that there was a rather large variability in the overall potentiation ( i . e . in time course and magnitude ) observed in recordings from individual cells , as demonstrated by the six typical cases of recording from different cells reported in Fig 1C . As will be discussed later , this finding is important for a better understanding of the interplay among the different processes underlying the induction of plasticity at each synaptic contact . Additional properties of t-LTP are summarized in Table 1 and suggest that , in all cases , t-LTP induction was found to be postsynaptic and NMDA receptor-dependent . Instead , expression was found to be pre-synaptic for the 70x 1:1 protocol , post-synaptic for the 25x 1:4 protocol , and mixed for the 50x 1:2 protocol . The pre- or post-synaptic expression of t-LTP was experimentally determined by i ) analyzing synaptic responses to short latency ( 50 ms ) paired pulses inducing pre-synaptic short term plasticity ( i . e . paired-pulse facilitation ) , ii ) by infusing an inhibitor of AMPA receptor insertion into the postsynaptic membrane via the recording pipette solution , iii ) by testing the AMPA/NMDAR current ratio , and iv ) by using analysis of the coefficient of variation of EPSPs pre- vs . post LTP induction ( see Fig . 2 in [11] ) . E . g . pre-synaptic 1:1 t-LTP changes the glutamate release probability of release and therefore changes the temporal dynamics of short-term plasticity . Conversely , the post-synaptically expressed 1:4 t-LTP does not change short-term plasticity , but rather changes postsynaptic AMPA/NMDAR current ratio and depends on incorporation of new AMPA receptors into the postsynaptic membrane ( all respective data shown in Fig . 2 of [11] ) . Of note , the 1:1 t-LTP protocol was composed of 1 EPSP paired with 1 backpropagating action potential ( bAP ) , whereas the 1:4 t-LTP protocol was composed of 1 EPSP paired with 4 ( instead of 1 ) bAPs . Thus the 1:1 t-LTP protocol can be considered as being included ( i . e . being a part of ) in the 1:4 t-LTP protocol . One might thus expect that the mechanisms triggered by the 1:1 t-LTP protocol should also be activated by the 1:4 t-LTP protocol , but this was not experimentally observed [11] . Other experimental suggestions that could be used to further constrain the model implementation: ( i ) An increase in postsynaptic intracellular calcium , [Ca2+]i , was necessary to initiate a complex chain of biochemical reactions leading to the vesicular release of BDNF [3]; this process has stochastic dynamics that are ~10 times slower than glutamate release which resulted in a large variability of the time course of postsynaptic BDNF release with respect to the triggering event of a transient [Ca2+]i elevation [26] . ( ii ) The postsynaptic BDNF release could last from a few seconds up to approximately 300 s ( [20] , supplementary Fig . 5 in [11] ) . ( iii ) There is no 1:1 ( i . e . pre-synaptic ) t-LTP expressed following 1:4 t-LTP stimulation [11]; this result may imply the existence of an additional mechanism , triggered by the 1:4 t-LTP protocol which is able to block the induction of 1:1 t-LTP . In summary , these experimental observations form a useful set of properties that give specific indications on what the model must be able to reproduce to be considered a reasonable representation of the many biochemical pathways that can be involved . In agreement with experimental suggestions [22 , 23] , the model was based on the assumption that any given individual synaptic contact , following the appropriate conditioning protocol , will change its state in an all-or-none manner . This was an important point to consider in comparing model and experimental findings since experimental recordings are customarily carried out from the soma , whereas the stimulation most likely involved an unknown number of synapses located in a relatively wide range of distances from the soma . The progressive increase in synaptic potentiation over time may thus be the result of an ensemble dynamics where different synapses undergo potentiation at different times . Unless explicitly stated otherwise , in discussing the model implementation we will always refer to individual synapses . The biochemical pathways that we considered for this work are schematically represented in Fig 2A , and it is based on the hypothesis that distinct biochemical pathways are activated by different levels of intracellular [Ca2+] in the postsynaptic compartment [27 , 28] . In our model , there were three different [Ca2+]i thresholds , θ1 , θ2 , and θ3 . Ca2+ entry can independently occur through NMDA receptor or voltage-gated Ca2+ channels , both explicitly included in our model and known to drive postsynaptic BDNF secretion [29] . A transient [Ca2+]i increase above each threshold activated one or many pathways in the spine head . The [Ca2+]i range below θ1 corresponded to the non-plastic regime , i . e . any combination of pre- and/or post-synaptic input did not alter the current state of the pre- and/or post-synaptic mechanisms . Above θ1 , it activated the 1:1 t-LTP signaling cascade ( Fig 2A , blue boxes ) , which released a yet to be identified retrograde messenger ( RM ) . This release activated presynaptic processes ( RM proc and presyn proc in Fig 2A ) resulting in a persistent increase in stimulus induced presynaptic glutamate release ( release ) , in agreement with the change in the paired-pulse ratio observed experimentally [11] . Experimental findings suggest that neither nitric oxide ( NO ) [11] nor endocannabinoids were involved as RM . The fusion of postsynaptic BDNF vesicles was activated by a larger and more long-lasting Ca2+ transient ( [Ca2+]i>θ2 ) , which may be obtained with the 70x 1:4 t-LTP protocol ( Fig 2A , dark pink blocks ) . The largest Ca2+ transients ( [Ca2+]i>θ3 ) activated biochemical reactions blocking RM production . The rationale for this choice was that experimental recordings clearly show that the 1:4 t-LTP protocol did not induce presynaptic LTP [11] . For this to happen , there must be an activity-dependent ( postsynaptic ) process blocking the biochemical pathways leading to presynaptic LTP . In the model , we made the simple assumption that this process could be a Ca2+-dependent block of the retrograde messenger release , occurring for a [Ca2+]i threshold ( θ3 ) that is higher than the one for LTP induction . Other model behaviors were not affected by this assumption , and this scheme left open the possibility , for spines in which [Ca2+]i reaches an intermediate concentration ( θ2<[Ca2+]i< θ3 ) , to account for a t-LTP with mixed pre- and post-synaptic mechanisms of expression that is obtained with a 1:2 t-LTP protocol [11] . The complete set of kinetic equations implementing the model ( introduced in the next paragraph ) were included into the membrane equation for each of the 18 explicitly modeled spines ( see Methods ) . To roughly take into account the local dendritic temporal integration process , 12 of the 18 spines were distributed on one oblique dendrite ( Fig 2B ) , whereas the remaining 6 spines were distributed on a different dendrite ( see blue bracket in Fig 2B ) and had different values for the θi ( see Table 2 ) . The presynaptic mechanisms specific for this work were added to the phenomenological model discussed in [31] , and described by the following set of equations: dxdt=zτrec−USE⋅x⋅δ ( t−tspike ) ( 1 ) dydt=−yτin+USE⋅x⋅δ ( t−tspike ) ( 2 ) z=1−x−y ( 3 ) where δ ( t ) is a delta function , tspike was the time of arrival of a spike at the pre-synaptic terminal , the variables x , y , and z are the fraction of resources in the recovered , active , and inactive states , respectively , and USE was proportional to the glutamate released by each synaptic stimulation . They reproduced the stereotypical synaptic response dynamics between pyramidal neurons under physiological conditions . The values for the presynaptic parameters were those used in Ref . [31] , with USE0 = 0 . 1 , τrec = 0 . 8 sec , and τin = 3 ms . This presynaptic mechanism has been previously shown to reproduce experimental findings on the normalization of temporal summation of synaptic inputs targeting distal or proximal dendrites of CA1 pyramidal neurons [32] . USE was additionally modulated by retrograde messenger-dependent pathways described by the following equations: dRMdt=αRM⋅ ( RM−RMinf ) +αCRM⋅S ( cai , θ1 , σ1 ) ⋅[1−S ( cai , θ3 , σ3 ) ]−αRMp⋅ ( RM−RMinf ) ⋅S ( cai , θRM , σRM ) ( 4 ) dRMpdt=αRMp⋅ ( RM−RMinf ) ⋅S ( cai , θRM , σRM ) −αpp⋅RMp ( 5 ) dppdt=αpp⋅RMp ( 6 ) USE=USE0⋅ ( 1+αRMpU⋅S ( pp , θU , σU ) ) , ( 7 ) where RMp and pp were presynaptic processes activated in cascade by RM accumulation in the synaptic cleft ( “RM proc” and “presyn proc” in Fig 2A ) , and S ( i , j , k ) =11+e ( j−i ) /k is the typical sigmoidal logistic function ubiquitously observed in biological systems [33] . Our hypothesis is that the activation of these mechanisms follows a dose-response curve . These processes are usually implemented with a sigmoid or a Hill function . Although the latter can be more easily related to the biomolecular pathways it represents , it also implies a significantly higher computational cost ( for NEURON running on a PC we verified a 35% difference in CPU time ) . This occurs because of the internal representation of the computational algorithms used to calculate an exp ( in a sigmoid function ) or a power ( in the Hill function ) on any given computer . Since we plan to use this model on a large-scale network , we have preferred to implement these curves with a sigmoid function . The cai was the intracellular calcium concentration [Ca2+]i . Note that pp does not have a decay term . This ensures that a potentiated synapse does not spontaneously fall back to its non-potentiated state . It is technically possible to continue to present the induction protocol for infinite time yielding to RM release and consequent infinite growth of pp . However , this exploratory modeling work does not consider this remote possibility . Post-synaptic mechanisms are activated by different levels of [Ca2+]i , with the instantaneous Ca2+ dynamics determined by the complex interaction between AMPAR and NMDAR conductances , voltage-gated Ca2+ channels , and all other active and passive membrane properties . All the equations regulating the instantaneous Ca2+ dynamics were taken from a previously published CA1 neuron model [30] ( ModelDB a . n . 55035 ) . We reported here only the equations of the new mechanisms introduced in this work , and directly related to the synaptic transmission pathways using [Ca2+]i as an input ( see Methods for detail on how to access the full model ) . The fusion of BDNF-containing vesicle with the spine head membrane is a complex process , possibly involving several biochemical pathways for which there are not enough experimental constraints to build a detailed kinetic scheme . For this reason , we implemented the effective action of these pathways using two mechanisms , accounting for the dependency of BDNF vesicle fusion probability and delay with respect to the STDP induction protocol . The first mechanism is implemented with an empirical variable , which we called intracellular signaling ( is ) . It is based on the experimental findings [11] suggesting that the fusion of BDNF-containing vesicles occurs only for conditioning protocols consisting of at least 25 induction stimuli repeated at a frequency close to 0 . 5 Hz , while no fusion was achieved in response to test stimulations at 0 . 05 Hz . In the model , this was obtained by increasing is by a fixed amount every time [Ca2+]i crossed the θ2 threshold , and decreasing it with a time constant of 8 s . The fusion was allowed to occur only for is>0 . 15 . With the second mechanism , we took into account the experimental findings ( [20] , supplementary Fig . 5 in [11] ) showing that the fusion of an individual vesicle containing BDNF , when activated , is a stochastic process occurring over a relatively long time window . We modeled all the involved processes by assuming that the fusion process happened with probability pf ( defined for [Ca2+]i>θ2 ) , and delay df , calculated as: pf={[Ca]i−θ2[Ca]i_max−θ2 ( is>0 . 15 ) 0 ( is≤0 . 15 ) ( 8 ) df=300⋅ ( 1−[Ca]i−θ2[Ca]i_max−θ2 ) ⋅rand[0 , 1] , ( 9 ) and assuming that Fused_vescicles=f ( pf , df ) ( 10 ) The function f ( pf , df ) keeps track of how many vesicles , in each synapse , have fused with the plasmatic membrane and were in the process of releasing BDNF . For each synapse , this function is increased by 1 with probability pf after a time interval df from the instant at which [Ca2+]i crosses the θ2 threshold . The function is updated , asynchronously for each vesicle , every 1 ms of simulated time , theoretically leading to a minimum interval of 1 ms between the start of a new vesicular release , with a maximum number of available vesicles in each synapse set at 200 , consistent with experimental data [34] . The function decreases by 1 ( with a minimum value of zero ) every time a vesicle has been fused for 30min . Random numbers from a uniform distribution in the interval [0–1] were used to choose the values for df , and pf; it should be stressed that this choice should not be considered as parametric randomization but , rather , as a way to introduce into the model an intrinsic stochastic behaviour . During the time the [Ca2+]i remains above the θ2 threshold the process leading to the release of a quantal amount of BDNF is active . In this time window , the fusion process of individual vesicles is initiated with probability pf and results in an actual fusion starting at a random time df ( up to 300 sec ) from activation , in agreement with experimental observations ( [20] , supplementary Fig . 5 in [11] ) . This also means that for [Ca2+]i remaining for a prolonged time above threshold , more fusion processes are started . Once a vesicle has fused with the membrane , it continuously releases a fraction of the stored mBDNF and proBDNF for some time . The experimental evidence for this process is indirect , and it suggests a lower and an upper bound for the overall process: the release lasts for at least 5 min [29] , but the overall LTP induction proceeds for approximately 30 min [11] . We made the somewhat simplifying and minimal assumption that the BDNF release lasts for 30 min . However , if this assumption would be invalidated by new experimental data , for example with longer experimental recordings of the BDNF release from single vesicles , the model could be straightforwardly revised by including an additional variable activated by a short BDNF release and slowly decaying over a period of 30 min . In any case , it is important to stress that in order to be consistent with the available experimental findings , the process modulating the magnitude of induced LTP must have a time course of approximately 30 min . The ratio between mBDNF and proBDNF inside these vesicles is unknown . Indirect experimental evidence [35 , 36] indicate for the mBDNF:proBDNF proportion a value in the range of 10% to 90% . This ratio also depends on the pH inside the vesicle [20 , 37] . Since in our mouse brain slices we detected ~66% mBDNF vs . ~33% proBDNF in cell lysates , we used a 70%:30% proportion of mBDNF and proBDNF , respectively . In the Golgi apparatus and in BDNF-containing vesicles proBDNF can be cleaved by protein convertases ( PC ) into mBDNF and BDNF pro-peptide . Following the release , remaining proBDNF can be cleaved by plasmin or matrixmetallo proteinases [20] . To empirically model extrasynaptic diffusion and reuptake [38 , 39] , mBDNF , proBDNF , and PC were all assumed to decay at a constant rate αdiff . The overall level of mBDNF present in the synaptic cleft determined the extent of TrkB receptor activation , Through a chain of postsynaptic processes represented by postsyn in Fig 2A , TrkB induces t-LTP by increasing the AMPA receptor conductance [11] . We implemented these processes as: dproBDNFdt=αfuse⋅0 . 3⋅Fused_vesicles⋅v_BDNF−αPC⋅PC⋅proBDNF−αdiff⋅proBDNF ( 11 ) dmBDNFdt=αfuse⋅0 . 7⋅Fused_vesicles⋅v_BDNF+αPC⋅PC⋅proBDNF−αdiff⋅mBDNF ( 12 ) dPCdt=αfuse⋅Fused_vesicles⋅v_PC−αdiff⋅PC ( 13 ) TrkB=mBDNF⋅S ( mBDNF , θTrkB , σTrkB ) ( 14 ) dpostdt=αpost⋅TrkB ( 15 ) gAMPA=gmax⋅[1+αAMPA⋅S ( post , θAMPA , σAMPA ) ] , ( 16 ) where gAMPA is the peak AMPA conductance , gmax its maximum value before LTP , and post represents the long-term effects of TrkB-dependent processes on the overall AMPA conductance . Note that post does not have a decay term . This ensures that a potentiated synapse does not spontaneously fall back to its non-potentiated state . It is technically possible to continue to present the induction protocol for infinite time yielding to BDNF release and consequent infinite growth of post . However , this exploratory modeling work does not consider this remote possibility . The overall model was too complex to attempt an automatic fitting procedure , especially considering that there were not enough clear experimental constraints to reduce the number of free parameters . For this reason , the parameters were set in two steps: 1 ) for each block shown in Fig 2A , an initial estimate for the involved parameters was obtained by presenting inputs that mimic the signals that could be generated in the full model , and manually adjusting the values to obtain what we considered a reasonable output signal; 2 ) test simulations of the full model were carried out with all spines placed on the dendrites . In this latter step , which can take into account the non-linear interaction between a spine and a backpropagating action potential , the parameters were further adjusted in such a way to result in an overall LTP level consistent with the experimental findings shown in Fig 1A . It is important to stress that the key point in this paper was not to explore the parameter space or to find their best values but to study if , how , and to what extent , the proposed scheme was able to take into account the basic experimental findings on BDNF-dependent spike-time-dependent LTP . As mentioned when discussing Fig 2 , we explicitly modeled eighteen independent spines , each containing the mechanisms described above with the parameters reported in Table 2 . To introduce the physiological variability of the biochemical pathway dynamics in the model , the αpp value in each synapse was drawn from a random uniform distribution . The number of synapses was not important for the scope of the paper . We found it a convenient number to illustrate and demonstrate that the overall effect measured at the soma was the result of a number of independent synapses . The key point here is that , as we will discuss later , the experimental findings cannot be reproduced by modeling a single synapse or a group of identical synapses . It should also be noted that there are many sources of noise that could affect the model behaviour . For example , random background synaptic activity could jitter the interaction between the elicited EPSPs and the bAPs . However , due to the large number of stimuli repetitions and the slow processes that they activate , this contributed to the overall behaviour in a way similar to the random localisation of the spines . The same would be with variability in the morphological and/or electrophysiological spine parameters . The table shows only the model parameters introduced in this work . All model files and the Python scripts used to run the simulations described in the paper are available for public download under the ModelDB section of the Senselab database ( http://senselab . med . yale . edu , a . n . HYPERLINK "http://modeldb . yale . edu/244412" 244412 ) . In summary , we have introduced a biophysical model of spike timing-dependent LTP at the Schaffer collateral synapse s of hippocampal CA1 pyramidal neurons . The model took explicitly into account , for the first time , several experimental findings on the BDNF-dependent biochemical pathways . In Fig 3A , we plotted the membrane potential at a spine head during a conditioning stimulus in which a synaptic activation ( arrow ) was paired with a bAP elicited with a Δt = +5 or +50 ms . The same time course was typically observed at all synapses . Note that for Δt = +50 ms ( Fig 3A , thin grey trace ) the synaptic activation and the bAP could be considered as completely separate events . In this case , the maximum voltage deflection observed in the spine head was approximately 22 mV during the EPSP alone and 17 mV for the bAP . With a Δt = +5 ms , the two events overlapped and summed nonlinearly , with a maximum deflection of 66 mV . The nonlinear summation of an EPSP paired with a properly timed bAP has been experimentally observed [40] , and in our model was a key factor in inducing LTP . It can be explained by considering that the depolarization caused by the synaptic activation has the effect of inactivating the KA channels , allowing a bAP arriving within a relatively narrow time window to better propagate in the dendrite and the spine . The resulting depolarization released the NMDA receptor Mg-block and allowed a supralinear Ca2+ influx . The [Ca2+]i time course , recorded in the 12 spines distributed along one of the oblique dendritic branches during a synchronous activation of all synapses , is shown in Fig 3B ( colored lines correspond to different spines ) . For comparison , we also plotted [Ca2+]i in a single spine for Δt = +50 ms ( grey trace ) . In all 12 spines of the dendritic segment shown at higher magnification in Fig 2B , the [Ca2+]i transiently raised above the θ1 threshold for a Δt = +5 ms , whereas none of the spines in the other branch ( compare blue bracket in Fig 2B ) reached the θ1 threshold ( remaining coloured transients in Fig 3B ) . The [Ca2+]i transient was significantly different among spines . This occurred because the back-propagation of an AP depends on the local dendritic properties . Since the RM release is proportional to the amount of [Ca2+]i above the θ1 threshold , the spines with larger [Ca2+]i transients ( e . g . bright green and red traces in Fig 3B ) were able to accumulate in a shorter time the amount of RM required to activate the pre-synaptic mechanisms of plasticity . This resulted in t-LTP induction ( in terms of an increase in the glutamate release ) earlier than in other spines ( Fig 3C dark green and brown traces ) . Spines for which there was a higher release of RM were potentiated earlier and with a faster transition ( bright green trace ) ; spines with lower RM release switched to a potentiated state later and with a slower transition ( e . g . dark green trace ) . Two spines did not release a sufficient amount of RM to trigger potentiation ( Fig 3B cyan and thick grey ) . In only one spine the [Ca2+]i transients crossed also the θ2 threshold triggering the postsynaptic potentiation mechanisms with a time course depending on TrkB activation ( Fig 3D bright green trace ) . In agreement with the experiments ( Fig 1A ) , it took around 25 min after the induction protocol to switch all synapses to a potentiated state . In the model , we assumed that this could be caused by the slow time constants of the biochemical pathways involved with the retrograde messenger ( see αpp in Table 2 ) . As expected , since the 1:1 t-LTP protocol was in general not able to generate enough Ca2+ entry to cross the θ2 threshold , the AMPA conductance , which was modulated by the post-synaptic plasticity mechanisms , did not increase for all but one of the synapses ( Fig 3D ) . Taken together these results suggest that , in order to be consistent with the experimental findings , it was necessary to make the physiologically reasonable assumption that the RM-dependent mechanisms needed to generate a different response at each synaptic location , which is a physiologically plausible condition . This was an important issue that is usually not considered in implementing subcellular models for synaptic transmission . Alternatively , it is possible that long time constants in downstream processes ( not explicitly modeled here ) , such as the incorporation of new glutamate-containing vesicles into the readily releasable pool , are responsible for the approximately 25 min delay in completing the induction of synaptic potentiation [41 , 42] . Pairing one synaptic stimulation with four bAPs ( Fig 4A ) resulted in [Ca2+]i transients spanning a range covering all thresholds ( Fig 4B both panels ) . For 2 of the 12 spines in one branch , the [Ca2+]i transient crossed only the θ1 threshold , resulting in a pre-synaptic t-LTP induction ( Fig 4C , cyan and thick grey traces ) . In other 2 synapses ( Fig 4B left panel , light brown and dark green traces ) it was θ2<[Ca2+]i<θ3 . This indicated the activation of both pre- and post-synaptic mechanisms . For the other 10 synapses , the [Ca2+]i crossed also the θ3 threshold , eliciting the activation of all the post-synaptic ( but not presynaptic ) pathways , with a consequent long-term potentiation of the AMPA peak conductance ( Fig 4D ) . For the 6 spines in the other dendritic branch , [Ca2+]i crossed the θ3 threshold in all spines , but only two spines were potentiated ( Fig 4D , yellow and magenta traces ) , while the other 4 spines were not because they did not release and accumulate enough BDNF in the cleft to activate the downstream signaling cascade . This behaviour allowed us to point out a suggestion of our model that will turn out to be extremely important later: BDNF release was necessary but not sufficient to trigger postsynaptic t-LTP . The model suggested that BDNF must accumulate in the synaptic cleft up to an amount sufficient to activate TrkB receptors , i . e . the release must be sufficiently frequent and strong . During the 25 stimulus repetitions , this condition was achieved for only two of the 6 spines ( Fig 4D , yellow and magenta traces ) . These findings show how the effect of a 25x 1:4 t-LTP conditioning , as observed from the soma , could result from a complex dendritic signal integration process independently occurring in each synapse , and involving diverse biochemical pathways that may interplay in different ways . The results described above were discussed in terms of the processes occurring at the single spine level . We now turn our attention to the average results obtained from several cells . In preliminary simulations , we found that synaptic location and biochemical dynamics in individual spines may result in quite different temporal profiles for the observed t-LTP . This suggests that the overall time course was the result of t-LTP induction and expression mechanisms at each synapse , which respond to the same conditioning protocol by switching to a potentiated state at different times . During a manual trial and error procedure , we thus found a possible combination of synaptic potentiation times that best represented the average experimental results ( Fig 5A , compare circles with open squares ) . The model results for the average EPSP slope measured at the soma were in the range of the experimental data for different values of Δt ( see Fig 5B , red and blue squares ) . The model was also able to reproduce several experimental recordings obtained from individual cells , as shown in Fig 5C for three examples using the 1:1 or 1:4 protocol . For these cases , we found that in order to match the recordings from an individual cell , it was sufficient to distribute the spines along the branch and assume different values for αUSE and αgampa . This is physiologically plausible since these parameters represent the amount of presynaptic glutamate released in the cleft and the density of postsynaptic AMPA receptors , respectively . These factors and the location of activated spines along the dendrite can be expected to be quite different among cells . These results show that the set of pathways included in the model is able to take into account the main mechanisms underlying BDNF-dependent t-LTP in hippocampal CA1 pyramidal neurons . The model can , therefore , be conveniently used to investigate additional BDNF-dependent effects at these synapses . To test the behavior of the model under different conditions of BDNF release , we selected two physiologically plausible cases related to specific biochemical processes that may be of particular interest . As mentioned before , the proportion of mBDNF and proBDNF inside the BDNF-containing vesicles is currently unknown , and it may be an important factor in regulating the series of biochemical cascades that they activate . Based on indirect measurements [38 , 43] , in our control model we had set this proportion to 70:30 ( see the postsynaptic mechanisms section ) . To test what would happen if this ratio was changed for physiological or pathological reasons , we carried out a set of simulations using the 25x 1:4 t-LTP protocol using a 30:70 proportion for mBDNF:proBDNF . The simulation results are summarized in Fig 6A ( green squares ) . As expected , lowering the initial amount of mBDNF ( Fig 6B from control , solid red line , to mBDNF30 , solid green line ) resulted in a weaker t-LTP in the first minutes after induction ( see open green squares in Fig 6A up to 6 min ) . In Fig 6B we show that in this case the mBDNF level was partially restored to the level observed with a 70:30 ratio ( compare the red dashed line with the solid green line ) by the cleaving action of PC on the higher level of proBDNF . The lower concentration of mBDNF yielded a higher latency in the expression of potentiation ( see green squares in Fig 6A in the range of 8–16 min after induction of t-LTP ) and an overall t-LTP after 30 min that was approximately 50% less pronounced than in control . This may appear surprising since it could be argued that the overall amount of BDNF released was the same as in control . This effect can be explained by the lower overall mBDNF accumulation during the 25 min period after conditioning , caused by extrasynaptic diffusion and reuptake processes . The delay in the accumulation of mBDNF resulted in a lower total concentration in the cleft at 25 min from induction failing to potentiate the spines with least [Ca2+]i influx . The final result was a failure to potentiate those synapses in which the BDNF accumulation was not enough to activate TrkB receptors and trigger the potentiation of the relative AMPA conductance . It should be stressed that here we were interested in illustrating the possible consequences of changing the mBDNF:proBDNF ratio before they were released from the post-synaptic vesicles . An additional effect , not included in this model , would be caused by the two factors preferentially binding , after their release , to different receptors ( proBDNF to p75 receptors versus mBDNF binding TrkB receptors ) followed by receptor mediated endocytosis ( discussed e . g . in [3] ) . Another process that we investigated was the duration for the vesicular release of BDNF . In contrast with the extremely fast glutamatergic mechanisms of release , the processes underlying BDNF release are known to last for minutes ( [11 , 26 , 29]; reviewed e . g . in [20] ) . To be consistent with experimental findings , under control conditions we used a 30 min time window , but what would be the consequence of having the release reduced or compressed into a shorter time window ? For example , reducing the release to a 15 min window corresponds to a proportionally reduced BDNF level; the model predicted that this would result in a 40% reduction in t-LTP expression , with respect to control ( Fig 6A , open black squares ) . Instead , compressing ( i . e . releasing the same amount of BDNF during a shorter time interval ) the entire release process leads to an overall faster dynamics , through which the same amount of BDNF was released within a shorter time; under this condition the model predicted that , although the final amount of t-LTP would be the same as control ( Fig 6A , blue circles ) , the maximum induction would be reached much earlier . The results obtained with the model were rather robust against changes in starting parameters ( see Fig A in S1 Text ) , and suggesting that the overall time course of t-LTP recorded at the soma could reflect the compound effect of the dynamics of the release process at the individual synaptic contacts . The kinetic scheme of the model was constructed to include a small set of building blocks needed to take into account the constraints that can be derived from experimental findings on the biochemical pathways that may be involved in BDNF-dependent t-LTP . We considered it a useful template to explore the consequences of selective changes in one or more of the pathways , in the attempt to elaborate a few model predictions fostering future experimental work . We were particularly interested in figuring out the possible mechanisms that could be exploited to rescue normal cognitive functions that under pathological conditions reduce synaptic transmission , such as Alzheimer’s disease ( AD ) . For example , we could assume that under AD conditions AMPA/NMDA receptors were less efficient in providing the depolarization needed to activate enough Ca2+ entry and to activate LTP induction [44–47] . This could occur , for example , because of AD-dependent alterations in the process of binding glutamate to receptors or in a conformational change affecting their peak conductance [44] . A possible rescue mechanism , in this case , could be a minimal perturbation of the pathways , in such a way to increase the magnitude of LTP that could be induced by the STDP protocol . Experimental findings suggest two possible ways in which this could be obtained . The first way is related to the clear separation in the induction processes between the pre- and post-synaptic t-LTP pathways . It has been shown that both processes did not occlude each other [11] . In the model , we have obtained this condition by including in the kinetic scheme a block of retrograde messenger production for high levels of [Ca2+]i . By removing this block , one could obtain a further increase in potentiation ( with a 1:4 protocol ) , since the two pathways would both be activated . In Fig 6C , we showed the results after removing the “RM block” from the kinetic scheme ( see Fig 2A ) . Under this condition , the combined pre- and post-synaptic t-LTP expression during a 1:4 protocol would be substantially higher than in control , with an overall 50% increase ( Fig 6C closed red squares ) . It should be stressed that , to the best of our knowledge , this was the only explanation for the surprising experimental finding that a 1:4 t-LTP protocol does not activate a presynaptic component that is instead induced by a 1:1 t-LTP protocol . A second possibility to increase the LTP level during pathological conditions may be to increase BDNF release . There are clear experimental indications suggesting that BDNF expression , and therefore most likely also BDNF release , is increased in response to physical exercise ( [48–53]; for a recent review see [54] ) . A direct connection between exercise , elevated brain BDNF levels , and rescue of synaptic function in Alzheimer’s disease has been described recently [55] . Also , a compensatory elevation of BDNF in AD affected brain areas has been reported ( reviewed e . g . in [56 , 57] ) , indicating that BDNF-dependent compensation of AD related synaptic deficits might be a spontaneously occurring endogenous protective mechanism that exists even in the absence of physical exercise—but that is further exploited—by physical exercise . Next , a set of simulations was then carried out with the BDNF release being increased in all spines by 20% ( Fig 6C , closed cyan squares ) . Also under these conditions , the model predicted an increase of the overall potentiation , which in this case was 25% higher than control . However , it should be noted that this increase would strongly depend on the number of spines that were not potentiated by the BDNF released under control conditions . An increase in BDNF release would thus recruit also those synapses and could result in an increase of the overall potentiation observed at the soma . A specific increase of BDNF release , without affecting other processes , has been previously demonstrated experimentally [11] . Taken together , the model predictions might contribute to design experimental investigations aiming to enhance LTP induction in order to rescue neuronal pathologies underlying learning deficits under pathological conditions .
BDNF signaling drives a widespread arsenal of synaptic functional and structural plasticity processes that shape synaptic circuits in many regions of the mammalian brain . Failure of this signaling pathway is known to take part in the progressive development of pathological neurodegenerative diseases such as AD [58] and Huntington's disease [59 , 60] with loss of memory formation and recall . To begin to sort out the role and intricacies of the many biochemical pathways involved in these processes , a detailed kinetic model of the BDNF signaling driving LTP would be very useful . The main aim of this study was thus the construction of such a model , making available to the scientific community a starting modular representation , able to capture most of the current knowledge on BDNF-driven t-LTP . The minimal set of signaling pathways implemented here , constrained by specific experimental findings [11] , include a cascade of reactions that can be individually subjected to further additions/extensions to take into account more specific molecular interactions . It can be argued that the process of releasing BDNF that will act on the same compartment where the release occurred is rather peculiar . However , it should be noted that this type of autocrine signaling is often observed for hormones and other growth factors . For BDNF , this is discussed in [3] , and most recent experimental evidence is demonstrated in [11 , 13 , 61 , 62] . The proposed kinetic scheme is a useful framework that can be extended in future work to explore in more details additional mechanisms that are known to be involved in synaptic plasticity , such as Arc protein-dependent processes [63] , the role of proBDNF/p75 signaling in t-LTD , or CAMKII auto-phosphorylation to account for synaptic bi-stability [19] and its role in both learning and forgetting [64–65] . The model provides an explicit biophysically and physiologically plausible representation , at the subcellular level , of the interplay among the timing of evoked pre- and post-synaptic activity , the active properties of the membrane , and the intracellular Ca2+ dynamics . The study of the conditions under which a local Ca2+ influx can trigger the induction of t-LTP in Schaffer collateral to CA1 neurons , allowed us to suggest the requirements for BDNF-dependent signaling at individual synaptic contacts . The results obtained in this work go beyond a simple reproduction of the main experimental results on the BDNF-dependent processes underlying t-LTP . We were able to make experimentally testable predictions on how and to what extent it would be possible to affect the overall magnitude of t-LTP , by manipulating specific synaptic transmission pathways . In this respect , we explored three possible alterations: Another important result of the model is the clear evidence that , what is observed with electrophysiological recording at the soma , cannot be reproduced from scratch by modeling a single synaptic contact and/or a single point neuron . We consider this a key point for both modelers and experimentalists: modelers , interested in implementing biologically accurate rules for synaptic transmission , plasticity , and dendritic signal integration , must take into account the physiological variability of the biochemical processes , independently occurring at individual synapses according to the local electrophysiological and biochemical environment [72]; experimentalists , interested in extracting as much as possible information from somatic recordings , will better understand the large scatter in the time course of STDP induced synaptic changes at individual synaptic contacts , and could eventually use the model to deduce location and biochemical properties of the potentiated synapses from the STDP time course . The mechanisms we combined to reproduce the cascade of gating activations could configure a highly unstable dynamical system . In order to demonstrate that our model is robust against small changes of the parameters , we performed 5 new simulation sets , each of which including two simulations where a specific model parameter was increased or decreased by a small fraction . These parameters were chosen because they control the most critical points of the dynamical system . For both the BDNF- and the RM-mediated plasticity chain , we have chosen to vary the [Ca2+] threshold that controls the activation of the mechanism and the gain factor that controls the amount of induced potentiation . In our model , these parameters control the beginning and the end of the chain of reactions leading to synaptic plasticity . Therefore , if the model would introduce instability at any point of its cascade of reactions , this would become evident after changing the [Ca2+] threshold and not after changing the final gain factor . In contrast , if instability was present in the system , but it was due to non-linear interaction of the potentiated synapses , the unstable behavior would be triggered by changes in either of the two parameters . For the RM-mediated plasticity chain , we selected an additional parameter in the middle of the chain . This controls the RM production and makes it relevant to distinguish between pre- or post-synaptic location of the potentially instable mechanism . The modeling results are shown in the Fig A in S1 Text . The panels A and B show the effects of parameter perturbations in the 25x 1:4 t-LTP-driven , BDNF-mediated plasticity chain . It is evident that a 20% change of the gain factor αAMPA ( Eq 16 ) effective at the end of the activation chain yields an effect similar to a 10% change in the [Ca2+] threshold θ2 ( Eqs 8 and 9 ) , which affects the entire cascade of gating activations . The panels C and D show results for the same kind of test for the RM mediated 70x 1:1 T-plascticity chain RM-mediated . Also in this case , a parametric perturbation at the beginning ( θ1 Eq 4 ) , in the middle ( θRM Eq 5 ) , and at the end ( αRMpU Eq 7 ) of the cascade of mechanisms yielded similar effects . These results indicate that the model is loosely sensitive to small perturbations of its key parameter , the [Ca2+] threshold . In the model construction , we used a sharp transition for the [Ca2+] threshold θ1 having set σ1 = 0 . 01e-3 mM to simplify the model tuning . In order to show that this very specific choice is irrelevant with respect to our results , we performed one supplemental simulation using a 100 times smoother transition ( σ1 = 0 . 001 mM = θ1/46 , Eq 4 ) . However , this simulation did not disclose any changes in the model behavior . One limitation of the model is the lack of LTD pathways in the kinetic scheme . The reason for this choice is the current lack of sufficiently detailed experimental constraints on the BDNF-dependent pathways that could be involved in the induction of LTD . The model can be considered a relatively simple template that can be extended to include LTD mechanisms , as soon as more experimental data become available . It should be noted that a synaptic stimulation after postsynaptic spikes ( a classic induction protocol for spike-time-dependent LTD ) will not generate enough Ca2+ entry to activate the cascade of processes leading to t-LTP ( Fig B in S1 Text ) . Finally , this model provides a foundation to investigate the multi-scale link between the short t-LTP induction period ( in the seconds range ) and those cases in which plastic changes occur at a much longer time-scale ( several minutes ) , such as BDNF-dependent STDP . The delay in BDNF release may encode for crossing a threshold of relevance caused by a stimulation that shall be remembered for many seconds . For example , this delay can take into account the time of arrival of rewards expected in procedural learning and mediated by dopamine signals . So far there is experimental evidence for BDNF vesicular release lasting at least 5 min [11 , 29] . The release of BDNF beyond this limit , if confirmed by experimental recordings , opens up the possibility to further fine-tune , on a longer time scale , the extent of how much of the initial learning stimulus is finally converted into a long-lasting memory . These processes might involve additional biochemical cascades regulating neuromodulatory transmitter signaling such as dopamine , noradrenaline or acetylcholine pathways [72] . The overall organization of these mechanisms can thus provide , for example , the neural correlate for the synaptic eligibility traces expected in reinforcement learning to solve the distal reward problem [73] . The model presented here can be used to investigate this type of problem , first at the subcellular level in a single cell and then to extract an effective and computationally more efficient algorithm to be used in large-scale network simulations , where the use of the full implementation would require prohibitively long computing times .
STDP experiments were performed on transversal hippocampal slices ( 350- to 400-μm thickness ) from either P15-P23 Wistar rats ( Charles River , Sulzfeld , both sexes ) , or P25-P35-day-old male BDNF+/- mice bred on a C57Bl/6J genetic background [74] or their WT littermates , respectively , as described previously ( compare [11] and references therein ) . All electrophysiological experimental data were taken from [11] . In short , timing-dependent ( t- ) LTP was induced with repeated pairings of one presynaptically induced EPSP , evoked by stimulation of Schaffer collaterals and one , two , or four postsynaptic APs induced by somatic current injection ( 2–3 ms , 1 nA ) via the recording electrode in current clamp configuration . T-LTP was induced after a 10 min baseline recording . Cells were held in the current clamp mode at -70 mV . Pairings were repeated 20–150 times depending on the specific protocol . T-LTP was induced by pre-post pairings ( indicated by positive spike timings ) consisting of either 1 EPSP/1 AP stimulation ( 70–100 repeats at 0 . 5 Hz ) , 1 EPSP/2 AP stimulation ( 50 repeats at 0 . 5 Hz ) , or 1 EPSP/4 AP stimulation ( 20–35 repeats at 0 . 5 Hz ) . Whole-cell recordings were performed at 30 . 5 °C ± 0 . 2 °C , with pipettes ( pipette resistance 6–10 MΩ ) filled with internal solution containing ( in mM ) : 115 potassium gluconate , 10 HEPES , 20 KCl , 4 Mg-ATP , 0 . 3 Na-GTP , 10 Na-phosphocreatine , 0 . 00075 CaCl2; pH was adjusted to 7 . 4 using KOH ( 280–290 mosmol/kg ) . The bath solution contained ( in mM ) : 125 NaCl , 2 . 5 KCl , 25 NaHCO3 , 0 . 8 mM NaH2PO4 , 20 glucose , 2 CaCl2 , 1 MgCl2 , saturated with 95% O2 and 5% CO2 ( pH 7 . 4; 304–306 mosmol/kg ) . Whole-cell recordings were obtained using either an EPC8 patch clamp amplifier connected to a LiH8+8 interface or an EPC10 patch clamp amplifier ( HEKA , Germany ) operated with PATCHMASTER software ( HEKA , Germany ) . Data were filtered at 3 kHz and digitized at 10 kHz . Data analysis was performed using FITMASTER ( HEKA , Germany ) and Mini Analysis software ( Synaptosoft , USA ) . Data analysis was performed as described previously [11] . Matched unpaired controls ( negative controls ) were performed for quality control with ongoing stimulation of 45min but in absence of any t-LTP induction paradigm [11] . Every recording of t-LTP started with 10 min baseline recording of EPSP at 0 . 05 Hz . EPSP slopes were calculated from the initial 2 ms after EPSP onset . The mean slope of this baseline was set to 100% . After the 10 min of baseline recording ( time interval from -10 to 0 min in all graphs ) the STDP protocol was executed , and all subsequent EPSP slopes were normalized to the 100% value during baseline recording . To describe the experimental results , we extensively use the terms induction and expression to refer to two distinct phases of t-LTP; LTP induction is the pattern of electrical activity in pre- and postsynaptic neurons that triggers the second messenger processes ( e . g . intracellular Ca2+ or cAMP elevation ) that set in motion the biochemical processes underlying enhanced synaptic transmission; LTP expression is the biochemical mechanism that accounts for the altered synaptic strength at the potentiated synapse , e . g . incorporation of new postsynaptic AMPA receptors , increased probability of presynaptic transmitter release , spine growth . All simulations were implemented using the NEURON simulation environment ( v7 . 4 , [75] ) using the variable time step feature . Additional custom code was written in Python . Model files will be available for public download under the ModelDB section of the Senselab database ( http://senselab . med . yale . edu , accession number 244412 ) . For all simulations , we started from a pre-existing CA1 pyramidal cell model [30] , downloaded from Senselab ( accession number 55035 ) . In this model , already validated against a number of different experimental findings achieved in rat CA1 neurons [30] , voltage-gated sodium ( INa ) and delayed rectifier potassium ( KDR ) conductances were uniformly distributed throughout the dendrites , while the KA [30] and Ih ( hyperpolarization induced inward current ) conductance linearly increased up to 500 μm from the soma . For the purposes of this paper , the peak KA conductance was increased by 30% , with respect to its original value to account for the shorter dendrite of CA1 neurons in mice compared to rats . To take into account local dendritic integration processes and allow for some physiological variability in the subcellular mechanisms underlying plasticity at individual synapses , eighteen explicit dendritic spines were modeled . The differences among spines are described in Results . Each spine was implemented with two compartments , one for the spine neck and one for the spine head . Following experimental indications [76] , the spine neck compartment was 2 μm long with a diameter of 0 . 5 μm , and the spine head was 0 . 264 μm long and 1 μm thick . Active ion channels were inserted in the spine head and included L- , N- , and T-type Ca2+ ion channels and Ca2+-dependent K channels downloaded from a previously published model ( [77] , ModelDB acc . n . 151126 ) . The resting Ca2+ concentration was set at 50 nM and , consistently with experimental evidences for the spine specific Ca2+ dynamics [76] , the spine intracellular [Ca2+] extrusion pump and buffering mechanisms were approximated with a single exponential decay ( τCa = 12 ms [76] ) . AMPA and NMDA receptor channels were placed on the spine head . Their kinetic models were adapted from [78] to include the vesicular cycling dynamics of glutamate release [31] . The effects of glutamate reuptake and diffusion away from the synaptic cleft were modeled using a fast exponential decay ( τglu = 0 . 1 ms ) for the extracellular glutamate concentration [79 , 80] . With respect to the original model [78] , we did not use the slow adapting and [Ca2+]i-dependent component of the AMPA kinetic , which in the original paper take into account a special form of BCM like synaptic meta-plasticity rule . Also , in our simulations the intracellular and extracellular concentration of Na+ and K+ ions was fixed . For the purposes of this work , we considered the following stimulation protocols delivered at 0 . 5Hz: In all cases , simulations were repeated using different delays , Δt from +5 ms to +30 ms , between the pre- and post-synaptic activation . Post-synaptic action potential activation was obtained with a suprathreshold somatic current pulse ( 1 nA for 2 . 5 ms ) . Test stimuli were delivered at 0 . 05 Hz before and after the induction protocols described above . The EPSP elicited by the last test stimulus delivered before initiation of the t-LTP induction protocol was used to compute the reference ( 100% ) EPSP slope ( maximum value of the time derivative in the 2 ms after EPSP onset ) . All figures show the EPSP slopes computed from the EPSP elicited after t-LTP induction and normalized by the reference slope described above . | Storing memory traces in the brain is essential for learning and memory formation , and it occurs through synaptic plasticity processes . Timing-dependent Long-Term Potentiation ( t-LTP ) is a physiologically relevant type of synaptic plasticity that results from the repeated sequential firing of action potentials ( APs ) in pre- and postsynaptic neurons . T-LTP is observed during learning in vivo and is a cellular correlate of memory formation . T-LTP can be elicited by different patterns of combined pre- and postsynaptic activity that recruit distinct synaptic growth processes underlying t-LTP . The protein Brain-Derived Neurotrophic Factor ( BDNF ) is released at synapses and mediates synaptic plasticity in response to specific patterns of t-LTP stimulation in the theta frequency band , while other patterns mediate BDNF-independent t-LTP . Here , we developed a realistic computational model that accounts for our previously published experimental results of BDNF-independent 1:1 t-LTP ( 70 repeats of pairing 1 presynaptic with 1 postsynaptic AP ) and BDNF-dependent 1:4 t-LTP ( 25 repeats of pairing 1 presynaptic with 4 postsynaptic APs ) . The model explains the magnitude and time course of both t-LTP forms and allows predicting t-LTP properties that result from altered BDNF turnover . Since BDNF levels are decreased in demented patients , understanding the function of BDNF in memory processes is important to counteract neurodegenerative diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"neurochemistry",
"vesicles",
"nervous",
"system",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"synaptic",
"plasticity",
"neurotransmitters",
"cellular",
"structures",
"and",
"organelles",
"neuronal",
"dendrites",
"excitatory",
"postsynaptic",
"potentials",
"developmental",
"neuroscience",
"animal",
"cells",
"glutamate",
"biochemistry",
"cellular",
"neuroscience",
"anatomy",
"synapses",
"cell",
"biology",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neurophysiology"
] | 2019 | A kinetic model for Brain-Derived Neurotrophic Factor mediated spike timing-dependent LTP |
The genetic component of complex disease risk in humans remains largely unexplained . A corollary is that the allelic spectrum of genetic variants contributing to complex disease risk is unknown . Theoretical models that relate population genetic processes to the maintenance of genetic variation for quantitative traits may suggest profitable avenues for future experimental design . Here we use forward simulation to model a genomic region evolving under a balance between recurrent deleterious mutation and Gaussian stabilizing selection . We consider multiple genetic and demographic models , and several different methods for identifying genomic regions harboring variants associated with complex disease risk . We demonstrate that the model of gene action , relating genotype to phenotype , has a qualitative effect on several relevant aspects of the population genetic architecture of a complex trait . In particular , the genetic model impacts genetic variance component partitioning across the allele frequency spectrum and the power of statistical tests . Models with partial recessivity closely match the minor allele frequency distribution of significant hits from empirical genome-wide association studies without requiring homozygous effect sizes to be small . We highlight a particular gene-based model of incomplete recessivity that is appealing from first principles . Under that model , deleterious mutations in a genomic region partially fail to complement one another . This model of gene-based recessivity predicts the empirically observed inconsistency between twin and SNP based estimated of dominance heritability . Furthermore , this model predicts considerable levels of unexplained variance associated with intralocus epistasis . Our results suggest a need for improved statistical tools for region based genetic association and heritability estimation .
Risk for complex diseases in humans , such as diabetes and hypertension , is highly heritable yet the causal DNA sequence variants responsible for that risk remain largely unknown . Genome-wide association studies ( GWAS ) have found many genetic markers associated with disease risk [1] . However , follow-up studies have shown that these markers explain only a small portion of the total heritability for most traits [2 , 3] . There are many hypotheses which attempt to explain the ‘missing heritability’ problem [2–5] . Genetic variance due to epistatic or gene-by-environment interactions is difficult to identify statistically because of , among other reasons , increased multiple hypothesis testing burden [6 , 7] , and could artificially inflate estimates of broad-sense heritability [8] . Well-tagged intermediate frequency variants may not reach genome-wide significance in an association study if they have smaller effect sizes [9 , 10] . One appealing verbal hypothesis for this ‘missing heritability’ is that there are rare causal alleles of large effect that are difficult to detect [4 , 11 , 12] . These hypotheses are not mutually exclusive , and it is probable that a combination of models will be needed to explain all heritable disease risk [13] . The standard GWAS attempts to identify genetic polymorphisms that differ in frequency between cases and controls . A complementary approach is to estimate the heritability explained by genotyped ( and imputed ) markers ( SNPs ) under different population sampling schemes [14 , 15] . Stratifying markers by minor allele frequency ( MAF ) prior to performing SNP-based heritability estimation allows the partitioning of genetic variation across the allele frequency spectrum to be estimated [16] , which is an important summary of the genetic architecture of a complex trait [16–23] . This approach has inferred a contribution of rare alleles to genetic variance in both human height and body mass index ( BMI ) [16] , consistent with theoretical work showing that rare alleles will have large effect sizes if fitness effects and trait effects are correlated [18 , 20–25] . Yet , simulations of causal loci harboring multiple rare variants with large additive effects predict an excess of low-frequency significant markers relative to empirical findings [4 , 26] . SNP-based heritability estimates have concluded that there is little missing heritability for height and BMI , and that the causal loci simply have effect sizes that are too small to reach genome-wide significance under current GWAS sample sizes [14 , 16] . Further , extensions to these methods decompose genetic variance into additive and dominance components and find that dominance variance is approximately one fifth of the additive genetic variance on average across seventy-nine complex traits [27] . When taken into account together with results from GWAS , these observations can be interpreted as evidence that the genetic architecture of human traits is best-explained by a model of small additive effects . However , a recent large twin study found a substantial contribution of dominance variance for fourteen out of eighteen traits [28] . The reason for this discrepancy in results remains unclear . One possibility is a statistical artifact; for example , twin studies may be prone to mistakenly infer non-additive effects when none exist . Another possibility , which we return to later , is that this apparently contradictory results are expected under a different model of gene action . The design , analysis , and interpretation of GWAS are heavily influenced by the “standard model” of quantitative genetics [29] . This model assigns an effect size to a mutant allele , but formally makes no concrete statement regarding the molecular nature of the allele . Early applications of this model to the problem of human complex traits include Risch’s work on the power to detect causal mutations [30 , 31] and Pritchard’s work showing that rare alleles under purifying selection may contribute to heritable variation in complex traits [17] . When applied to molecular data , such as SNP genotypes in a GWAS , these models treat the SNPs themselves as the loci of interest . For example , influential power studies informing the design of GWAS assign effect sizes directly to SNPs and assume Risch’s model of multiplicative epistasis [32] . Similarly , the single-marker logistic regression used as the primary analysis of GWAS data typically assumes an additive or recessive model at the level of individual SNPs [33] . Finally , recent methods designed to estimate the heritability of a trait explained by genotyped markers assigns additive and dominance effects directly to SNPs [14 , 16 , 27 , 34] . Naturally , the results of such analyses are interpreted in light of the assumed model of gene action . A weakness of the multiplicative epistasis model [30 , 31] when applied to SNPs is that the concept of a gene , defined as a physical region where loss-of-function mutations have the same phenotype [35] , is lost . Specifically , under the standard model , the genetic concept of a failure to complement is a property of SNPs and not “gene regions” ( see [36] for a detailed discussion of this issue ) . We have recently introduced an alternative model of gene action , one in which risk mutations are unconditionally deleterious and fail to complement at the level of a “gene region” [36] . This model , influenced by the standard operational definition of a gene [35] , gives rise to the sort of allelic heterogeneity typically observed for human Mendelian diseases [37] , and to a distribution of GWAS “hit” minor allele frequencies [4 , 26] consistent with empirical results [36] . In this article , we explore this “gene-based” model under more complex demographic scenarios as well as its properties with respect to the estimation of variance components using SNP-based approaches [34] and twin studies . We also compare this model to the standard models of strictly additive co-dominant effects , and multiplicative epistasis with dominance . We further explore the power of several association tests to detect a causal gene region under each genetic and demographic model . We find significant heterogeneity in the performance of burden tests [36 , 38 , 39] across models of the trait and demographic history . We find that population expansion reduces the power to detect causal gene-regions due to an increase in rare variation , in agreement with work by [22 , 23] . The behavior of the tests under different models provides us with insight as to the circumstances in which each test is best suited . In total , our results show that modeling gene action is key to modeling GWAS , and thus plays an important role in both the design and interpretation of such studies . Further , the model of gene-based recessivity best explains the differences between estimates of additive and dominance variance components from SNP-based methods [27] and from twin studies [28] and is consistent with the distribution of frequencies of significant associations in GWAS [4 , 26] . Further , the genetic model plays a much more important role than the demographic model , which is expected based on previous work on additive models showing that the genetic load is approximately unaffected by changes in population size over time , [21 , 22] . Consistent with recent work by [23] , we find that rapid population growth in the recent past increases the contribution of rare variants to total genetic variance . However , we show here that different models of gene action are qualitatively different with respect to the partitioning of genetic variance across the allele frequency spectrum . We also show that these conclusions hold under the more complex demographic models that have been proposed for human populations [21 , 40] .
As in [36] , we simulate a 100 kilobase region of human genome , contributing to a complex disease phenotype and fitness . The region evolves forward in time subject to neutral and deleterious mutation , recombination , selection , and drift . To perform genetic association and heritability estimation studies in silico , we need to impose a trait onto simulated individuals . In doing so , we introduce strong assumptions about the molecular underpinnings of a trait and its evolutionary context . How does the molecular genetic basis of a trait under natural selection influence population genetic signatures in the genome ? This question is very broad , and therefore it was necessary to restrict ourselves to a small subset of molecular and evolutionary scenarios . We analyzed a set of approaches to modeling a single gene region experiencing recurrent unconditionally-deleterious mutation contributing to a quantitative trait subject to Gaussian stabilizing selection . The expected fitness effect of a mutation is always deleterious because trait effects are sampled from an exponential distribution . Therefore , we do not allow for compensatory mutations that may occur in more general models of stabilizing selection . Specifically , we studied three different genetic models and two different demographic models , holding the fitness model as a constant . Parameters are briefly described in Table 1 . We implemented three disease-trait models of the phenotypic form P = G + E . G is the genetic component , and E = N ( 0 , σ e 2 ) is the environmental noise expressed as a Gaussian random variable with mean 0 and variance σ e 2 . In this context , σ e 2 should be thought of as both the contribution from the environment and from the remaining genetic variance at loci in linkage equilibrium with the simulated 100kb region . The genetic models are named the additive co-dominant ( AC ) model , multiplicative recessive ( Mult . recessive; MR ) model and the gene-based recessive ( GBR ) model . The MR model has a parameter , h , that controls the degree of recessivity; we call this model the complete MR ( cMR ) when h = 0 and the incomplete MR ( iMR ) when 0 ≤ h ≤ 1 . Here , h = 1 corresponds to co-dominance , which is different from the typical formulation used when modeling the fitness effects of mutations directly . It is also important to note that here recessivity is being defined in terms of phenotypic effects; this may be unusual for those more accustomed to dealing directly with recessivity for fitness effects . An idealized relationship between dominance for fitness effects and trait effects of a mutation on an unaffected genetic background is shown in S15 Fig . The critical conceptual difference between recessive models is whether dominance is a property of a locus ( nucleotide/SNP ) in a gene or the gene overall . Mathematically , this amounts to whether one first determines diploid genotypes at sites ( and then multiplies across sites to get a total genetic effect ) or calculates a score for each haplotype ( the maternal and paternal alleles ) . For completely co-dominant models , this distinction is irrelevant , however for a model with arbitrary dominance one needs to be more specific . As an example , imagine a compound heterozygote for two biallelic loci , i . e . genotype Ab/aB . In the case of traditional multiplicative recessivity the compound heterozygote is wild type for both loci and therefore wild-type over all; this implies that these loci are in different genes ( or independent functional units of the same gene ) because the mutations are complementary . However , in the case of gene-based recessivity [36] , neither haplotype is wild-type and so the individual is not wild-type; the failure of mutant alleles to complement defines these loci as being in the same gene [35] . For a diploid with mi causative mutations on the ith haplotype , we may define the additive model as G A C = ∑ i = 1 2 ∑ j = 1 m i c i , j , ( 1 ) where ci , j is the effect size of the jth mutation on the ith haplotype . Each ci , j is sampled from an exponential distribution with mean of λ , to reflect unconditionally deleterious mutation . In other words , when a new mutation arises its effect c is drawn from an exponential distribution , and remains constant throughout its entire sojourn in the population . The GBR model is the geometric mean of the sum of effect sizes on each haplotype [36] . We sum the causal mutation effects on each allele ( paternal and maternal ) to obtain a haplotype score . We then take the square root of the product of the haplotype scores to determine the total genetic value of the diploid . G G B R = ∑ j = 1 m 1 c 1 , j × ∑ j = 1 m 2 c 2 , j ( 2 ) Finally , the MR model depends on the number of positions for which a diploid is heterozygous ( mAa ) or homozygous ( maa ) for causative mutations , G M R = ∏ j = 1 m A a ( 1 + h c j ) ∏ j = 1 m a a ( 1 + 2 c j ) - 1 . ( 3 ) Thus , h = 0 is a model of multiplicative epistasis with complete recessivity ( cMR ) , and h = 1 closely approximates the additive model when effect sizes are small . Here , phenotypes are subject to Gaussian stabilizing selection with an optimum at zero and standard deviation of σs = 1 such that the fitness , w , of a diploid is proportional to a Gaussian function [41] . w = e - P 2 2 σ s 2 ( 4 ) The AC and MR models draw no distinction between a “mutation” and a “gene” ( as discussed in [36] ) . The GBR is also a recessive model , but recessivity is at the level of a haplotype ( or allele ) and is not an inherent property of individual mutations ( see [36] for motivation of this model ) . Viewed in light of the traditional AC and MR models , the recessivity of a site in the GBR model is a function of the local genetic background on which it is found . Based on several qualitative comparisons we find that the GBR model is approximated by iMR models with 0 . 1 ≤ h ≤ 0 . 25 . However , no specific iMR model seems to match well in all aspects . The demographic models are that of a constant sized population ( no growth ) and rapid population expansion ( growth ) . The use of the MR model is inspired by Risch’s work [30 , 31] , linking a classic evolutionary model of multiple loci interacting multiplicatively [42 , 43] to the the genetic epidemiological parameter relative risk . Risch and Merikangas [44] used this model to calculate the power to detect causal risk variants as a function of their frequency and effect size . Pritchard extended Risch’s model to consider a trait explicitly as a product of the evolutionary process [17] . Pritchard’s work demonstrated that the equilibrium frequency distribution suggested an important role for rare deleterious mutations when a trait evolves in a constant sized , randomly mating population with recurrent mutation and constant effect sizes . However , multiplicative epistasis is only one model of gene action . Exploring the effect of different genotype-to-phenotype models on the population and quantitative genetic properties of complex traits is the focus of the current work . The amount of narrow sense heritability , h2 = ( VA ) / ( VP ) , explained by variants across the frequency spectrum is directly related to the effect sizes of those variants [29] . Thus , this measure is an important predictor of statistical power of GWAS and should inform decisions about study design and analysis [45] . Empirically , SNP-based estimates of heritability have inferred negligible dominance variance underlying most quantitative traits [27] . We have a particular interest in the amount of additive variance , VA , that is due to rare alleles and how much of genetic variance , VG , is attributable to VA under different recessive models . We follow the approach of [21] , by calculating the cumulative percent of VG explained by the additive effects of variants less than or equal to frequency x , ( VA;q ≤ x ) / ( VG ) . The product of this ratio and broad-sense heritability is an estimate of the narrow-sense heritability , h2 . This calculation is a population-wide equivalent to a SNP-based estimate of heritability in a population sample . In addition we calculate the same distribution for dominance effects ( VD;q ≤ x ) / ( VG ) using the orthogonal model of [27] . Methods based on summing effect sizes [29] or the site frequency spectrum [21] would not apply to the GBR model , because the effect of a variant is not independent of other variants ( e . g . , there is intralocus epistasis ) . Therefore , we resort to a regression-based approach , where we regress the genotypes of the population onto the total genetic value as defined in our disease trait models ( see Material and Methods ) . In the limit of Hardy-Weinberg and linkage equilibrium , the regression estimates are equivalent to standard quantitative genetic estimates [29] ( S14 Fig ) . For consistency , we applied the regression approach to all models . Overall , these distributions are substantially different across genetic models , demographic scenarios and model parameters ( Fig 1 ) . Under the AC model , all of VG is explained by additive effects if all variants are included in the calculation; in Fig 1 the solid variance curves reach unity in the AC panel . Low frequency and rare variants ( q < 0 . 01 ) explain a large portion of narrow sense heritability ( 26%–95% ) even in models without rapid population expansion . Further , the variance explained at any given frequency threshold increases asymptotically to unity as a function of increasing λ ( S4 Fig ) . While the total heritability of a trait in the population is generally insensitive to population size changes ( S1 Fig , see also [21 , 22 , 46] ) , rapid population growth increases the fraction of additive genetic variation due to rare alleles ( Fig 1 ) . Here , increasing λ corresponds to stronger selection against causative mutations , due to their increased average effect size . Recent work by Zuk et al . [24] , takes a similar approach and relates the allele frequency distribution directly to design of studies for detecting the role of rare variants . However , our findings contrast with those of Zuk et al . [24] and agree with those of Lohmueller [22] , in that we predict that population expansion will substantially increase the heritability , or portion of genetic variance , that is due to rare variants . Our results under the AC model agree with those of Simons et al . [21] , in that we find that increasing strength of selection , increasing λ in our work , increases the contribution to heritability of rare variants . However , under the GBR model and the cMR model the distribution of genetic variance over risk allele frequency as function λ is non-monotonic ( Fig 1 and S4 Fig ) . For all recessive models , we find that total VA is less than VG ( Fig 1 ) . For the MR models , all additional genetic variation is explained by the dominance variance component; in Fig 1 the dotted variance curves reach unity in the MR panels . As expected , genetic variation under the MR model with partial recessivity ( h = 0 . 25 ) is primarily additive [29 , 47] , whereas VG under the cMR model ( h = 0 ) is primarily due to dominance . The GBR model shows little dominance variance and is the only model considered here for which the total VG explained by VA+VD is less than the true VG for all λ . This can be clearly seen in Fig 1 where the dotted curves do not reach unity in the GBR panel . These observations concerning the GBR model are consistent with the finding of [27] that dominance effects of SNPs do not contribute significantly to the heritability for complex traits . Under the GBR model , large trait values are usually due to compound heterozygote genotypes ( e . g . , Ab/aB , where A and B represent different sites in the same gene ) [36] . Therefore , the recessivity is at the level of the gene region while the typical approach to estimating VA and VD assigns effect sizes and dominance to individual mutations . Thus , compound heterozygosity , which is commonly observed for Mendelian diseases ( see [36] and references therein ) would be interpreted as variation due to interactions ( epistasis ) between risk variants . Importantly , the GBR model assumes that such interactions should be local , occurring amongst causal mutations in the same locus . While the GBR model is reflective of the original definition of a gene in which recessive mutations fail to complement , we emphasize that this does not imply that mutations are necessarily exomic . The GBR model is of a general genomic region in which mutations act locally in cis to disrupt the function of that region with respect to a phenotype . The increase in the number of rare alleles due to population growth is a well established theoretical and empirical result [48–61] . The exact relationship between rare alleles [4 , 17 , 26 , 62 , 63] , and the demographic and/or selective scenarios from which they arose [21 , 22 , 64] , and the genetic architecture of common complex diseases in humans is an active area of research . An important parameter dictating the relationships between demography , natural selection , and complex disease risk is the degree of correlation between a variants effect on the disease trait and its effect on fitness [18 , 20–22] . In our simulations , we do not impose an explicit degree of correlation between the phenotypic and fitness effects of a variant . Rather , this correlation is context dependent , varying according to the current genetic burden of the population , the genetic background in which the variant is present and random environmental noise . However , if we re-parameterized our model in terms of [18] , then we would have τ ≤ 0 . 5 ( Gaussian function is greater than or equal to its quadratic approximation ) , which is consistent with recent attempts at estimating that parameter [20 , 65] . Our approach is reflective of weak selection acting directly on the complex disease phenotype , but the degree to which selection acts on genotype is an outcome of the model . While the recent demographic history has little effect on key mean values such as broad-sense heritability of a trait or population genetic burden ( S1 and S3 Figs ) , the structure of the individual components in the population which add up to those mean values varies considerably . The specific predictions with respect to the composition of the populations varies drastically across different modeling approaches . It is therefore necessary to carefully consider the structure of a genetic model in a simulation study . The conclusions reached here also hold when we consider more complex demographic scenarios relevant to human populations . Under the demographic model for European populations from [40] , the additive and GBR models show the same behavior as in Fig 1 ( S17 Fig ) . At all key time points where population size changes , VA = VG for the additive model , and the variance explained by rare mutations depends primarily on λ ( S17 Fig ) . For the GBR model , VA < VG ( as in Fig 1 ) , and plateaus at the same ratio VA/VG for all time points except immediately after the bottleneck , which results in a short-lived increase in VA/VG that is undetectable by the time growth begins ( S17 Fig ) . All recessive models ( GBR , iMR and cMR ) may show a transient increase in total VG after the bottleneck , depending on the value of λ ( S18 Fig ) . However , the GBR and iMR models with h > 0 . 25 showed a return to constant population size levels by the final time point . The changes in VA and VG under recessive models is likely due to the transfer of non-additive variation into VA during a bottleneck , which has been studied thoroughly in the theoretical literature [66 , 67] . As in Fig 1 , the genetic model , and not the demographic details , drive the relationship between mutation frequency and additive genetic variance . In agreement with existing literature , site based recessive models show complex dynamics during bottlenecks and population expansion ( S18 and S19 Figs ) . However , with respect to load , the GBR model behaves more like a codominant model and is largely insensitive to changes in population size ( S18 and S19 Figs ) . Thus , complex traits evolving under the GBR model are not expected to show large differences in load between extant human populations . The previous section shows that the relationship between genetic variance and allele frequency in the entire population strongly depends on the genetic model . Recent estimates of variance components from large population samples of unrelated individuals have inferred that dominance variance ( VD ) is negligible for most traits [27] . However , a recent study of more than 104 Swedish twins and 18 traits obtained a contradictory result , inferring significant non-additive variance for most traits , which was interpreted as VD [68] . In this section , we show that this apparent inconsistency is expected under certain models of gene action . We applied GREMLd , MAF-stratified GREMLd ( MS-GREMLd ) , and MAF-stratified Haseman-Elston regression ( see Methods for details ) . We found MS-GREMLd to be numerically unstable on our simulated data , and thus we present results for non-MS-stratified GREMLd . The numerical stability issues likely resulted from some combination of small number of SNPs per region ( O ( 1000 ) ) , low total VG in a region , or high variance in effect sizes across causal mutations [69] . Further , for large λ , where VG is primarily due to rare alleles ( Fig 1 ) , heritability in a sample may not reflect heritability in the entire population ( S13 Fig ) . Fig 2 shows the GREMLd additive and dominance heritability estimates , as compared to the respective population value , over λ . Under the cMR model ( h = 0 ) , the dominance component is much larger than the additive component as predicted from Fig 1 . When GREMLd is performed on cMR model data after removing variants with MAF ≤ 0 . 01 , as done in [27] , the total heritability estimate ( AD ) is quite accurate until λ ≥ 0 . 25 where a downward bias is observed . As anticipated , GREMLd using unfiltered data yields results with a slight upward bias [70] . However , for the iMR ( h = 0 . 25 ) model the filtered GREMLd estimates are only accurate for λ < 0 . 1 reflecting the preponderance of rare causal variants for larger values of λ . Unfiltered GREMLd estimates under the iMR ( h = 0 . 25 ) model show a slight upward bias for small values of λ , but are otherwise accurate . This shows that GREMLd is performing as expected under the site-based model for which it is designed . The MS-HE regression results are generally consistent with the GREMLd results . The GREMLd and MS-HE estimates are accurate under the GBR model when λ is small , because most heritability is additive in that case ( Fig 1 ) . However , under the GBR model , both filtered and unfiltered GREMLd heritability estimates show downward bias when λ is large ( Fig 2 ) . The MS-HE regression results reveal a similar pattern , which indicates that the downward bias for large values of λ is not strictly due to removal of rare variants in the filtered GREMLd analysis . Instead , the bias shown for large values of λ is likely due to the presence of substantial non-additive heritability , which is not captured by the dominance effects of SNPs . In contrast to the variance component methods , our simulated large twin studies provide approximately unbiased estimates of total heritability for large values of λ , but were biased upward for small effect sizes under the AC and GBR models ( Fig 2 ) . The variance in twin-study estimates was quite large , possibly because only a single locus was simulated rather than the whole genome . Formally , twin studies estimate an additive and a non-additive component of variance and interpreting the non-additive component as epistatic or dominance variance is a matter of perspective . However , the GBR model is inspired by the definition of a gene as a physical region in which recessive mutations leading to the same phenotypic outcome fail to complement [35] , consistent with the allelic heterogeneity observed for human Mendelian disorders ( see [36] for further discussion ) . Thus , the model of recessivity at the level of the gene region is picked up as non-additive variance in twin studies , but missed by variance component methods ( GREML and HE regression ) because the dominance in the GBR model is due to Ab/aB ( compound heterozygotes ) genotypes rather than a/a genotypes ( homozygotes for a specific loss of function variant ) assumed by variance component methods . Thus the contradictory results of applying variance component methods [27] and analysis of large twin studies [68] in order to estimate VA and VD may be interpreted as evidence for a model of gene action such as the GBR , which may be viewed as either recessivity at the haplotype/gene level or intralocus epistasis at the level of causative mutations in a single gene region . Both interpretations are valid . The alternative explanation is that we must assert that one of the study designs is generating artifacts . Both demography and the model of gene action affect the degree to which rare variants contribute to the genetic architecture of a trait ( Fig 1 ) . However , the different mappings of genotype to phenotype from model to model make it difficult to predict a priori the outcomes of GWAS under each model . Therefore , we sought to explicitly examine the performance of statistical methods for GWAS under each genetic and demographic model . We assessed the power of a single marker logistic regression to detect the gene region by calculating the proportion of model replicates in which at least one variant reached genome wide significance at α ≤ 10−8 ( Fig 3A ) . The basic logistic regression is equivalent to testing for association under the AC model . We simulated both a perfect “genotyping chip” ( all markers with MAF ≥ 0 . 05 ) and complete re-sequencing including all markers ( Fig 3B ) . One of the most prominent feature of Fig 3 is the curvature of power as a function of λ . This reflects the competing forces of increasing average genetic effect and decreasing average allele frequency which occurs as λ increases ( S5 Fig ) . As λ increases , the total genetic variance explained by the locus increases until the model enters the House-of-cards [71] regime . At which point , the genetic variance is much less dependent on λ ( S1 Fig ) . When λ is large , however , the average allele frequency does continue to decrease ( S5 Fig ) which drives power down . Across all genetic models , the single marker logistic regression has less power under population expansion ( Fig 3A ) . The loss of power is attributable to a combination of rapid growth resulting in an excess of rare variants overall [48–61] , and the increasing efficacy of selection against causal variants in growing populations [21] . While complete resequencing is more powerful than a gene-chip design , the relative power gained is modest under growth ( Fig 3A ) . Region-based rare variant association tests behave similarly with respect to population growth ( Fig 3B ) . There are important differences in the behavior of the examined statistical methods across genetic models . We focus first on the single marker tests ( Fig 3A ) . For gene-chip strategies , power increases for “site-based” models as recessivity of risk variants increases ( compare power for AC , iMR , and cMR models in Fig 3B ) . This increase in power is due to the well-known fact that recessive risk mutations are shielded from selection when rare ( due to being mostly present as heterozyogtes ) , thus reaching higher frequencies on average ( S5 Fig ) , and that the single-marker test is most powerful when risk variants are common [32] . Further , for the complete multiplicative-recessive model ( cMR ) , the majority of VG is due to common variants ( Fig 1 ) , explaining why resequencing does not increase power for this model ( Fig 3A ) . For single-marker tests , the GBR model predicts large gains in power under resequencing for intermediate λ ( the mean trait-effect size of newly arising causal mutations ) , similar to the AC or iMR model . But , when λ is larger power may actually be less under the GBR model than under AC or iMR . For all models , causal mutations are more rare with increasing λ ( S7 Fig ) . However , as a function of frequency , all VG may be attributed to VA or VD in the site-specific models whereas there is increasing intralocus epistasis in the GBR model as a function of λ ( Fig 1 ) . It is well-known that the single marker test has lower power when causal mutations have low frequencies , are poorly tagged by more common SNPs , or have small main effects [32 , 72] . Region-based rare variant association tests show many of the same patterns across genetic model and effect size distribution as single marker tests , but there are some interesting differences . The ESM test [36 , 73] is the most powerful method tested for the AC , iMR , and GBR models ( Fig 3b ) , with the c-Alpha test as a close second in some cases . For those models , the power of naive SKAT , linear kernel SKAT and SKAT-O , is always lower than the ESM and c-Alpha tests . This is peculiar since the c-Alpha test statistic is the same as the linear kernel SKAT test . The major difference between SKAT and ESM/c-Alpha is in the evaluation of statistical significance . SKAT uses an analytical approach to determine p-values while the ESM/c-Alpha tests use an explicit permutation approach . This implies that using permutation based p-values results in greater power . Yet , under the cMR model the linear kernel SKAT is the most powerful , followed by c-alpha . The cMR model does not predict a significant burden of rare alleles and so the default beta weights of SKAT are not appropriate , and the linear kernel is superior . The ESM test does poorly on this model because there are not many marginally significant low-frequency markers . It is logical to think that these tests would all perform better if all variants were included . The massive heterogeneity in the performance of region-based rare variant tests across models strongly suggests that multiple methods should be used when prior knowledge of underlying parameters is not available . In agreement with [22 , 74] , we predict that population growth reduces the power to associate variants in a causal gene region with disease status ( Fig 3 ) when the disease also impacts evolutionary fitness . We have recently released software to apply the ESM test to case control data [73] in order to facilitate applying this test to real data . It was noted by [4 , 26] , that an excess of rare significant hits , relative to empirical data , is predicted by AC models where large effect mutations contribute directly to fitness and the disease trait . We confirm that AC models are inconsistent with the empirical data ( Fig 4 ) , except when λ ≤ 0 . 01 . The empirical data in Fig 4 represent a pooled data set with the same diseases and quality filters as in [26] , but updated to include more recent data . The data are described in S1 Table , and can be visualized alone more clearly in S16 Fig . Close to half of the data comes from GWAS studies uploaded to the NHGRI database after 2011 , yet the same qualitative pattern is observed . This contradicts the hypothesis that the initial observation of an excess of common significant hits relative to the prediction under an AC model was simply due to small sample sizes and low marker density in early GWAS previously analyzed in [4 , 26] . Yet the initial observation is in fact robust and the meta-pattern provides an appropriate point of comparison when considering the compatibility of explicit population-genetic models with existing GWAS data . The GBR model predicts few rare significant hits and an approximately uniform distribution across the remainder of MAF domain ( Fig 4 ) , even for intermediate and large values of λ . For smaller values of λ , the GBR predicts an excess of common significant hits . The more uniform distribution of significant single markers seen under the GBR is consistent with the flatter distribution of genetic variance ( Fig 1 ) . If one considers trying to determine an approximate dominance coefficient in the GBR model , it would be found that there is a distribution of coefficients across sites . Yet , when simulating iMR model , we find that an intermediate degree of dominance , h = 0 . 25 , results in distribution of significant hits which is similar to the GBR results ( Fig 4 ) . Most of the models fail a KS test comparing the simulated and empirical distribution of significant hits ( S21 Fig ) . The cMR ( h = 0 ) model shows a visual excess of intermediate frequency variants ( Fig 4 ) , but this does not result in rejection under the KS test ( S21 Fig ) which is largely insensitive to deviations in the tails . According to the KS test , the remaining models ( AC , GBR , iMR ) perform best when there are fewer data points in the simulated data due to low GWAS power . This suggests that all models would be rejected with enough replicates . We note that there is no compelling reason to expect any specific value of λ to be a particularly good fit to the empirical data . The empirical data are composed of genome-wide data for multiple traits . We feel that the mutational parameters , λ and mutation rate to causal variants , are likely to vary across the genome and across traits . Thus , the empirical data reflect a mixture of different underlying models and ascertainment schemes . The reason we emphasize this feature of the data is to demonstrate that models with rare alleles of large effect do not necessarily imply a visual excess of rare significant GWAS hits . In consideration of the rare allele of large effect hypothesis , [62] proposed a model where multiple rare alleles dominate disease risk and create synthetic associations with common SNPs . However , later it was shown that this particular model was inconsistent with GWAS theoretically and empirically [4 , 26 , 75] . Here , we have shown that there exist models in which rare alleles explain a substantial portion of heritability that are not inconsistent with findings from GWAS . We find that the MAF distribution of significant hits in a GWAS varies widely with choice of genetic model . In particular , we confirm the results of Wray et al . [26] , that AC evolutionary models predict an excess of low frequency significant hits unless trait effect sizes are quite small . Also , the cMR model predicts an excess of intermediate and common significant hits . Utilizing a GBR model or an iMR model with h = 0 . 25−0 . 5 , reconciles this inconsistency by simultaneously predicting the importance of rare alleles of large effect and the correct allele frequency distribution among statistically significant single markers . Several empirical observations provide support for the presence of gene-based recessivity underlying variation for some complex traits in humans . The minor allele frequency distribution of significant GWAS hits is relatively flat [4 , 76] , which our results show is consistent with either the presence of small additive effect loci or gene-/site-based partially-recessive loci with intermediate to large effects ( Fig 4 ) . Models with loci of large additive effects predict an excess of rare significant hits . Oppositely , models with complete site-based recessivity predict an excess of common significant hits for all simulated mutation effect size distributions . SNP based estimates of dominance heritability are much lower than estimates of dominance from twins [27 , 68] . Of the models we explored , only the gene-based recessive model with intermediate to large effects is consistent with the difference between twin and SNP based estimates of dominance variance ( Fig 2 ) . Under a site-based recessive model of partial recessivity ( e . g . h = 0 . 25 ) , there should be no significant difference between estimates of dominance variance from SNP and twin studies , provided that the statistical assumptions are met for both approaches ( Fig 2 ) . These results are complementary to the work by Zuk et . al [24] , who show that twin studies can over estimate heritability under a model with gene interactions . It now appears clear that the underlying genetic model does not have the same impact on SNP-based and family based study designs; an issue which should be further explored . Our findings also support a more thorough investigation into the importance of compound heterozygosity in the genetics of complex traits . However , it may be difficult to directly observe non-additive gene-level effects through analysis of individual SNP markers . Additionally , the genetic model appears to be important in the design and analysis of association studies . While changes in population size do affect the relationship between effect size and mutation frequency [48–61] ( Fig 1 and S5 Fig ) , different mappings of genotype to trait value do this in radically different ways for the same demographic history ( Fig 1 ) . From an empirical perspective , our findings suggest that re-sequencing in large samples is likely the best way forward in the face of the allelic heterogeneity imposed by the presence of rare alleles of large effect . Resequencing of candidate genes [77–80] and exomes [40 , 81–86] in case-control panels have observed an abundance of rare variants associated with case status . Here we show that under a model of mutation-selection balance on the genic level , neither current single-marker nor popular multi-marker tests are especially powerful at detecting large genomic regions harboring multiple risk variants ( Fig 3 ) . However , we show that using permutations to derive p-values improves the power of SKAT [69] with a linear kernel ( equivalent test statistic to c-Alpha [38] ) . Similarly , another permutation based test , the ESM test [73] , has more robust power across demographic and genetic models ( Fig 3 ) . Conceptually , cis-effects arise naturally from the original definition of a gene in which mutant recessive alleles fail to complement [35] . We show that cis-effects within a locus , represented by the GBR model , can have an important impact on the population level architecture of a complex trait . This conclusion is important for future simulation studies as well as the interpretation of empirical data . It is important to note that despite our use of the term “gene-based” this model may apply to any functional genomic element in which there are multiple mutable sites affecting a trait in cis , not just to genes . From a theoretical perspective , our work motivates the development of a more generalized gene-based model to include arbitrary dominance and arbitrary locus size . Empirically , we find that the GBR model is broadly consistent with a variety of observations from the human statistical genetics literature . Thus , there is an evident need for improved region-based association tests and the development of genetic variance component methods for haplotypes .
Using the fwdpp template library v0 . 2 . 8 [87] , we implemented a forward in time individual-based simulation of a Wright-Fisher population with mutation under the infinitely many sites model [88] , recombination , and selection occurring each generation . We simulated populations of size N = 2e4 individuals for a time of 8N generations with a neutral mutation rate of μ = 0 . 00125 per gamete per generation and a per diploid per generation recombination rate of r = 0 . 00125 . Deleterious mutations occurred at a rate of μd = 0 . 1μ per gamete per generation . These parameters correspond to θ = 4Nμ = ρ = 4Nr = 100 and thus our simulation approximates a 100Kb region of the human genome . For simulations with growth , we simulated an additional 500 generations of exponential growth from Ni = 2e4 to Nfinal = 1e6 . This demographic model is much simpler than current models fit to empirical data [58] . However , this simple model allows us to more easily get a sense of the impact of population expansion [21 , 22] . 250 simulation trials were performed for each parameter/model combination unless specified otherwise . Broad-sense heritability can be calculated directly from our simulated data as H 2 = V G V P . We explored broad-sense heritability as a function of mean causative effect size λ under each model; λ ∈ {0 . 01 , 0 . 025 , 0 . 05 , 0 . 1 , 0 . 125 , 0 . 25 , 0 . 5} . We compare our simulation results to V G ∼ 4 μ d σ s 2 for additive models and V G ∼ 2 μ d σ s 2 for recessive models [71 , 89] . In our simulations , σ s 2 = 1 , and we tuned the environmental standard deviation σe to generate simulations for which E[H2] ∼ 0 . 04 or ∼ 0 . 08 . For E[H2] ∼ 0 . 04 , we set σe = 0 . 11 for the additive codominant model , σe = 0 . 075 for the gene based and complete multiplicative recessive models and σe = 0 . 098 for the incomplete mutliplicative recessive model ( h = 0 . 25 ) . For E[H2] ∼ 0 . 08 , we set σe = 0 . 075 for the additive codominant model , σe = 0 . 053 for the gene based and complete multiplicative recessive models and σe = 0 . 068 for the incomplete mutliplicative recessive model ( h = 0 . 25 ) . Genetic load is defined as the relative deviation in a populations fitness from the fitness optimum , L = ( w m a x - w ¯ ) / ( w m a x ) . We set the phenotypic optimum to be zero; Popt = 0 . When determining fitness for the site based models , we subtract one from all phenotypes . This implies that w m a x = e - P o p t 2 2 σ s 2 = 1 and that load is a simple function of the phenotypes of the population , L = 1 - e - P 2 2 σ s 2 . We also used the mean number of mutations per individual , and the mean frequency and effect sizes of segregating risk variants as proxies for the genetic load [21 , 90] . Lastly , we calculated Burden Ratios ( Br ) [91] as the ratio of load between an equilibrium and non-equilibrium population . We calculated Br using both the true load and the number of mutations per individual . We used an approach based on sequential ( type-1 ) regression sums of squares to estimate the contribution of the additive and dominance effects of variants to the total genetic variation due to a locus . Given a genotype matrix ( rows are individuals and columns are risk variants ) of ( 0 , 1 , or 2 ) copies of a risk allele ( e . g . all mutations affecting phenotype ) , we sort the columns by decreasing risk mutation frequency . Then , within frequency classes , columns were sorted by decreasing effect sizes . For each variant a dominance component was also coded as 0 , 2q , or 4q-2 according to the orthogonal model of [27] , where q is the frequency of the variant in the population . We then used the R package biglm[92] to regress the individual genetic values ( G in the previous section ) onto this matrix . The variance explained by the additive and dominance effects of the m markers with q ≤ x is then approximately r 2 = ( ∑ i = 1 m Σ S S r e g , i ) / ( S S t o t ) . Averaging results across replicates , this procedure results in a Monte-Carlo estimate of the fraction of VG that is due to additive and dominance effects of variants with population frequency less than or equal to x is ( VA;q ≤ x + VD;q ≤ x ) / ( VG;q ≤ 1 ) [21] . This fraction can be easily partitioned into strictly additive and dominance components . We employed three different SNP-based approaches to estimating heritability from population samples: GREMLd , minor allele frequency stratified GREMLd ( MS-GREMLd ) [27] , and MS-Haseman-Elston ( HE ) regression [93 , 94] . For comparison , we calculated the true total heritability in the sample as H s a m p l e 2 = ( V G ; s a m p l e ) / ( V P ; s a m p l e ) . Unfortunately , due to the nature of our simulated data MS-GREMLd did not result in sufficiently reliable results . Under MS-GREMLd , many replicates resulted in numerical errors in GCTA . These problems were present at a rate of less than 1/100 replicates using non-MS GREMLd , but were increased by splitting the data into multiple GRMs . Using raw individual phenotypic values as quantitative trait values , random samples from simulated populations ( n = 6000 ) were converted to . bed format using PLINK 1 . 90a [95] . PLINK was also used to test for HWE ( p < 1e−6 ) and filter on minor allele frequency . GCTA 1 . 24 . 4 [34] was used to make genetic relatedness matrices ( GRM ) for both additive and dominance components with the flags –autosome and –make-grm ( -d ) . For non-MS runs , we tested the effect of filtering on MAF by performing the analysis on unfiltered datasets and with markers with MAF < 0 . 01 removed . For MS estimates we stratified the additive and dominance GRM’s into two bins MAF ≤ 0 . 01 and MAF > 0 . 01 . GREMLd analysis was performed in GCTA with Fisher scoring , no variance component constraint and a max of 200 iterations . MS-HE regression was carried out by regressing the off diagonal elements of each GRM onto the cross product of the scaled and centered phenotypes in a multiple linear regression setting in R [96] . To simulate twin studies we sampled 2000 monozygotic ( MZ ) and 2000 dizygotic ( DZ ) twins pairs from the final generation of the simulations . Parents were sampled randomly without replacement . MZ twin pairs were formed by sampling a single gamete pair , one recombinant from each parent , and two environmental random deviates . DZ twin pairs were formed by sampling two gamete pairs , two recombinant gametes from each parent , and two environmental random deviates . Our simulated studies are ideal in that there are no correlated environmental effects , but potentially problematic due to low total heritability . We explored the use of structural equation modeling ( SEM ) using the package OpenMx [97] , but chose to rely strictly on estimates of twin correlation obtained directly from the data . For monozygotic ( MZ ) twins , we used only a single child gamete pair with two unique environmental deviates . For dizygotic ( DZ ) twins we used two child gamete pairs , each with a unique environmental deviate . Broad sense heritability is the correlation between MZ twin pairs; H2 = rMZ . Under a purely additive model , the DZ twin correlation should be half of the MZ twin correlation . Non-additive genetic components of phenotypic variance reduce the DZ twin correlation . If all non-additive heritability is due to dominance , then the dominance heritability can be calculated as twice the difference between the MZ twin correlation and two-times the DZ twin correlation: δ2 = 2* ( rMZ − 2*rDZ ) . The additive heritability can then be calculated as the difference between the broad-sense and non-additive component: h2 = H2 − δ2 = 4*rDZ − rMZ[29] . These direct estimates of MZ and DZ twin correlations in our simulations are reliable as we have no measurement error , shared environmental effects , gene-by-environment effects , or gene-by-gene interactions . Additionally , we only simulate a single genomic region contributing H2 ∼ 0 . 04 , which made use of SEM difficult numerically . This creates a limitation in that we can not discuss when a model with dominance is a better fit to the data than the additive only model . But , the benefit of using direct estimates is that we can clearly see what signals are present in the data . To further clarify the data visualization , we pooled our 512 twin-study replicates into groups of 8 , creating 64 sets of MZ-DZ twin phenotypes . This did not have an effect on the central tendencies of our estimates , but it reduced the variance . The twin study error bars in Fig 2 are based on 64 sets of 64 , 000 individuals , which is larger than a typical twin study . However , one reason our results have high variance is because we only simulate a single locus , rather than a whole trait . Following [36] , we sampled 3000 cases and 3000 controls from each simulated population . Cases were randomly sampled from the upper 15% of phenotypic values in the population , and controls were randomly sampled from within 0 . 5 standard deviations of the population mean ( as in [36] ) . This is the liability scale model ( see [29] ) . We define a “GWAS” to be a study including all markers with MAF ≥5% and a re-sequencing study to include all markers . In all cases we used a minor allele count logistic regression as the single marker test . For single marker tests , the p-value cut off for significance is p ≤ 1e − 08 which is common in current GWAS [62 , 98] . Power is determined by the percentage of simulation replicates in which at least one marker reaches genome wide significance . We applied multiple region-based tests to our simulated data , ESMK[36] , several variations of SKAT [39] and c-Alpha [38] . We used the R package from the SKAT authors to implement their test ( http://cran . r-project . org/web/packages/SKAT/index . html ) . The remaining tests were implemented in a custom R package ( see Software availability below ) . For the ESMK and c-Alpha we performed up to 2e6 permutations of case-control labels to determine empirical p-values . Common variants ( q ≥ 0 . 05 ) were removed prior to performing region-based rare variant association tests . Following [4 , 26] , we calculated the distribution of the minor allele frequency ( MAF ) of the most significant SNPs in a GWAS in empirical and simulated data . The empirical data was obtained from the NHGRI-EBI GWAS database ( http://www . ebi . ac . uk/gwas/ ) on 02/05/2015 . We considered the same diseases and applied the same filters as in Table 3 of [26] . Specific information regarding the empirical data can be obtained in S1 Table . In order to mimic ascertained SNP data , we sampled markers from our case/control panels according to their minor allele frequencies [99] , as done in [36] . Additionally , we removed all markers with MAF <0 . 01 to reflect common quality controls used in GWAS . The simulated data were grouped by genetic model , demographic scenario , heritability level , and mutation effect distribution . We then plotted the minor allele frequency of the most significant marker with a single-marker score −log10 ( p ) ≥8 , for all replicates where significant markers were present . Finally , we performed a two-sample KS test in R between each group of simulated GWAS hit allele frequencies and the empirical data . We simulated a demographic model for Europeans based on [40] as described in [21] . For simplicity , we ignored migration between the European ( EA ) and African American ( AA ) populations . The model was implemented using the Python package fwdpy version 0 . 0 . 4 , which uses fwdpp[87] version 0 . 5 . 1 as a C++ back-end . During the evolution of the EA population , we recorded the genetic variance in the population , VG , and the number of deleterious mutations per diploid ( a measure of genetic load [21] ) every 50 generations . In a separate set of simulations , we applied the regression method described above to calculate cumulative additive genetic variance as a function of allele frequency . Because the regressions are computationally demanding , we applied the method in the generation immediately before , and at the start of , any changes in population size . These simulations were run with no neutral mutations , and the recombination rate and mutation rate to causative mutations were the same as in the simulations described above . The Python scripts for these simulations and iPython/Jupyter notebooks used for generating figures are available online ( see Software availability section below ) . Our simulation code and code for downstream analyses are freely available at | Gene action determines how mutations affect phenotype . When placed in an evolutionary context , the details of the genotype-to-phenotype model can impact the maintenance of genetic variation for complex traits . Likewise , non-equilibrium demographic history may affect patterns of genetic variation . Here , we explore the impact of genetic model and population growth on distribution of genetic variance across the allele frequency spectrum underlying risk for a complex disease . Using forward-in-time population genetic simulations , we show that the genetic model has important impacts on the composition of variation for complex disease risk in a population . We explicitly simulate genome-wide association studies ( GWAS ) and perform heritability estimation on population samples . A particular model of gene-based partial recessivity , based on allelic non-complementation , aligns well with empirical results . This model is congruent with the dominance variance estimates from both SNPs and twins , and the minor allele frequency distribution of GWAS hits . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome-wide",
"association",
"studies",
"deletion",
"mutation",
"demography",
"population",
"genetics",
"twins",
"simulation",
"and",
"modeling",
"developmental",
"biology",
"mutation",
"genome",
"analysis",
"population",
"biology",
"research",
"and",
"analysis",
"methods",
"genetic",
"loci",
"people",
"and",
"places",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"evolutionary",
"biology",
"genetics",
"of",
"disease",
"computational",
"biology",
"human",
"genetics"
] | 2017 | A Model of Compound Heterozygous, Loss-of-Function Alleles Is Broadly Consistent with Observations from Complex-Disease GWAS Datasets |
Essential metabolic reactions are shaping constituents of metabolic networks , enabling viable and distinct phenotypes across diverse life forms . Here we analyse and compare modelling predictions of essential metabolic functions with experimental data and thereby identify core metabolic pathways in prokaryotes . Simulations of 15 manually curated genome-scale metabolic models were integrated with 36 large-scale gene essentiality datasets encompassing a wide variety of species of bacteria and archaea . Conservation of metabolic genes was estimated by analysing 79 representative genomes from all the branches of the prokaryotic tree of life . We find that essentiality patterns reflect phylogenetic relations both for modelling and experimental data , which correlate highly at the pathway level . Genes that are essential for several species tend to be highly conserved as opposed to non-essential genes which may be conserved or not . The tRNA-charging module is highlighted as ancestral and with high centrality in the networks , followed closely by cofactor metabolism , pointing to an early information processing system supplied by organic cofactors . The results , which point to model improvements and also indicate faults in the experimental data , should be relevant to the study of centrality in metabolic networks and ancient metabolism but also to metabolic engineering with prokaryotes .
Prokaryotes are the simplest contemporary life forms known , and nevertheless are characterized by an immense complexity . The debate on the features of such complexity and its breadth in the primordial life forms has been around for years [1–3] , and was furthermore expanded and detailed since the advent of systems biology [4–6] . The study of essential genes has been a crucial contribution for detangling this complexity , relating some proteins with cell viability in specific conditions [7] and others with cell viability in apparently all conditions [8 , 9] . Genome-wide essentiality studies based on collections of targeted mutants or generated by random mutagenesis have been conducted for a number of species , aiming mainly at antibiotic design or identifying industrially relevant targets [10–15] . These data have been made available in databases such as the Online Gene Essentiality Database ( OGEE ) [16] and the database of essential genes ( DEG ) [17] but their comparative and integrative analysis , although having already provided relevant insights , is still emerging . Early work comparing genome-scale essentiality data of Mycoplasma genitalium , Haemophilus influenza , Bacillus subtilis and Escherichia coli found that it is essentiality , not expressiveness , that drives gene strand bias [18] . Later , a critical review of genome-scale essentiality datasets conducted a preliminary analysis integrating six assays corresponding to four different species [19] . Functional differences were highlighted , as the smaller number of essential genes in flavin synthesis in B . subtilis , a species known to have an active riboflavin salvage capability . The authors of the DEG database have also conducted a pair of integrative analyses on large-scale essentiality data . The first concluded that there are less essential genes inside than outside genomic islands , and some of those are related with virulence [20] . The second study [21] added to a previous finding based only on E . coli essentiality data where it was proposed that essential genes are more evolutionarily conserved than non-essential genes [22] . Luo and others used the same type of analysis , based on synonymous and non-synonymous substitution rates for 23 bacterial species , to corroborate this finding [21] . The study indicates that the most evolutionarily conserved COG categories of essential genes are carbohydrate transport and metabolism; coenzyme transport and metabolism; transcription; translation , ribosomal structure and biogenesis; lipid transport and metabolism , and replication , recombination and repair . Genome-scale metabolic models ( GSMs ) are large curated repositories of metabolic data for individual species that expand possibilities of analysis of cellular physiology [23] . Apart from improving or suggesting new functional annotations by reconstructing whole pathways [24] , GSMs can be used for calculating metabolic fluxes that permit the prediction of , among others , lethal phenotypes [25] . Multi-species analysis of this type of phenotype predictions with different manually curated models has been scarce [26] , in part impaired by the poor knowledge basis for other species than the usual model organisms , but also by the deficient use of standards in building such models [27] . Comparative genomics is commonly used to find core essential genes for several species , and being based on the key evolutionary notion of orthology , to infer genes present in common ancestors [28] . Evolutionary parsimony indicates that genes present in a set of species have been vertically inherited from a common ancestor . Horizontal Gene Transfer ( HGT ) might have played a role even before the divergence of the three main domains [29] , but when a gene is present in all or most species of a phylogenetic branch , the most parsimonious scenario is vertical inheritance . In parallel with genetic comparisons , genome-scale metabolic models allow for functional comparisons at the level of metabolic capacities ( reactions ) . Building up on this methodological advantage , in this study , 36 experimental genome-scale essentiality assays were integrated with simulation results from 15 genome-scale metabolic models to reveal common patterns of essentiality . To this analysis the screening of full genome sequences of 79 prokaryotic species was added in order to find core conserved functions in prokaryotic biology . It is expected that this knowledge on the minimal metabolic functions of prokaryotic cells can not only help untangling the fundamental complexity of cellular systems but also , by building up on the concept of orthogonalization of metabolic modules [30] , here analysed in the form of metabolic subsystems ( pathways ) , improve future engineering approaches that use this type of organisms .
For all essentiality predictions performed in this study , 15 genome-scale metabolic models were chosen based on curation , validation , and comparability of the nomenclature of metabolites and reactions . These comprise 7 prokaryotic phyla , including one archaea . Ten of these models include more than 20% of the total of ORFs of the corresponding species . Table 1 summarizes the details on the models used , including species name , model ID , content and references . All GSMs were collected in SBML format and then parsed to model an environmental condition corresponding to rich media: all original exchange reactions in the model were set to a maximum uptake limit of -20 mmol gDW-1 h-1 to allow for the import of all transported metabolites ( including oxygen , whenever possible ) . Flux Balance Analysis ( FBA ) [46 , 47] was used to predict the essentiality of each metabolic reaction in all models . A threshold of 10% of the flux through the biomass reaction compared to the wild type was set as the limit to define an essential metabolic reaction . All modelling procedures were implemented in C++ and solved using IBM ILOG CPLEX solver . The Optflux platform [48] was used occasionally to benchmark results . For comparison of the essential reactions calculated for the 15 GSMs , some nomenclature inconsistencies were resolved: standardization of suffixes used in reaction IDs , removal of unnecessary or redundant indications of reversibility and species names allocated to reactions and other redundant tags . Irrelevant and irregular characters such as dashes were filtered out of the nomenclature ( the final list of standardized reaction ids for reactions essential at least once is provided in the S1 File ) . Large-scale experimental data on gene essentiality were collected from two databases , OGEE [16] and DEG [17] . The content of the databases was compared and DEG was chosen for it was considerably larger , including wider and clearer annotation metadata for 36 prokaryotic datasets ( Table 2 ) . Genes were mapped to the subsystems present in the latest Escherichia coli genome-scale metabolic model [49] using the DEG integrated nomenclature system of gene identifiers . All essential reactions obtained after GSMs analysis were also mapped according to this updated list of subsystems , using their standardized nomenclature ( see above; S2 File ) . To analyse the conservation and infer ancestry of all the genes annotated in metabolic subsystems of GSMs , a local protein blast was performed against representative genomes of all the 35 prokaryotic phyla with at least one fully sequenced quality genome in the NCBI genome database ( version June 2015 ) . For this task , translated genomes were selected and downloaded for 53 unique species of prokaryotes for which an available GSM could be found; to these , 26 representative genomes for phyla not modelled with GSMs were added . This totalled in 79 translated genomes representing the 35 fully sequenced phyla in the current tree of life of prokaryotes ( see S2 Fig , built with iTOL v . 4 . 2 . 1 [78] which can be reproduced in iTOL with the corresponding NCBI taxonomy ) . All of the protein-encoding genes of E . coli K12 ( RefSeq genome NC_000913 . 3 ) were used as queries and annotated to the subsystems of the latest E . coli model [49] . The threshold e-value considered was 1e-10 . All the procedures were implemented using the Biopython package [79] . For assessing the conservation of essential reactions and essential genes in each metabolic subsystem , the weighted sum of essentiality ( W ) was calculated for each subsystem m , as: Wm=∑i=1tni*i ( 1 ) ni being the number of reactions or genes of that subsystem essential in i models or datasets , where t is the total of models or datasets , 15 and 36 respectively . The score sum of experimental essentiality for each individual gene Sg was calculated as the number of times a gene was found essential ( E ) minus the number of times it was found not essential ( N ) in all experimental assays datasets: Sg=E−N ( 2 ) All statistical analyses were performed using R ( statistical software , version 3 . 1 ) . Hierarchical clustering was performed using the ‘pvclust’ R package [80] with binary distance as the dissimilarity metric and Ward 1 method as the linkage criterion . Pvclust was also used for assessing uncertainty by calculating approximately unbiased p-values via multiscale bootstrap resampling . Both the Fisher and Kolmogorov Smirnov tests were performed in R as well with the corresponding default parameters .
To analyse the validity of the essentiality results on a large scale , the different models were clustered based on single-reaction essentiality predictions and the different datasets available on DEG [17] were clustered based on the content of essential genes ( Fig 1 ) . In the case of the simulated essentiality ( Fig 1A ) , strongly supported clusters ( more than 80% of 1000 bootstrap replicas ) are phylogenetically consistent at the level of the phylum–they cluster in a statistically significant manner with at least one sister species–with the exception of the models of C . beijerinckii and P . putida . H . pylori and S . oneidensis show up in the same cluster , but not together with the rest of the Proteobacteria , but the two high-level clusters ( that exclude M . tuberculosis with a p-value of 98% ) are not statistically supported ( p-values of 58 and 62% for the left and right cluster respectively ) . The lower number of available exchange reactions in the models of H . pylori , S . oneidensis and P . putida ( 74 , 95 and 89 respectively ) compared with the models for other Proteobacteria ( K . pneumoniae , E . coli K12 , S . typhimurium and E . coli W with 289 , 299 , 305 and 310 respectively ) points to a justification for these results , as less exchanges cause more reactions in the network to be essential . C . beijerinckii’s model is also very restricted with regards to exchange reactions , with only 19 metabolic drains available . Regarding the experimental data ( Fig 1B ) , there is also a pattern of clustering taxonomically related species . One well-supported phylogenetic cluster is that of several gamma and beta-proteobacteria , including Acinetobacter baylyi , dataset II of E . coli K12 , three Salmonellas , one Shewanella and one Francisella . Others are the cluster of Bacteroidetes , the cluster with all three datasets of M . tuberculosis and the cluster of the alpha-proteobacteria , Sphingomonas and Caulobacter . The datasets of Pseudomonas aeruginosa PAO1 and Salmonella enterica subsp . Enterica serovar Typhimurium str . 14028S cluster together likely due to not being saturated genome-wide gene-essentiality screens ( these datasets are considerably smaller , with 117 and 105 genes respectively–see Table 2 for context ) . Surprisingly , Firmicutes are spread all across the tree . Also , both E . coli sets are very distant from each other . Although they were performed under rich media conditions , one yielded 609 essential genes and the other only 296 ( Table 2 ) . This difference is likely due to the use of different technologies to perform the large-scale assays , the first being random mutagenesis and the screening of mixed populations , and the second the screening of libraries of targeted mutants , as reviewed in [19] . Next , all the essential reactions calculated for the 15 GSMs were mapped to the corresponding metabolic subsystem ( see Methods; S1 and S2 Files ) . Different models show different proportions of essential reactions for each subsystem ( Fig 2 ) . For the majority of the models , the most essential subsystem is that of cofactor and prosthetic group biosynthesis . Nearly 48% of the essential reactions in the simulations with the GSM of E . coli K12–54 reactions–were related with this subsystem ( Fig 2 ) . Several of the predicted essential reactions were confirmed to be essential steps in the biosynthesis of the active forms of cofactors that cannot be directly uptaken , e . g . dihydrofolate synthase and dihydrofolate reductase for the biosynthesis of tetrahydrofolate and derivatives [81] , NAD kinase for obtaining NADP [82] and riboflavin synthase in some species [83] . For M . tuberculosis , D . ethenogenes , S . typhimurium and K . pneumoniae the most represented subsystems were glycerophospholipid or membrane lipids metabolism ( 30 . 5 , 25 . 1 , 21 . 8 and 29 . 2% of all essential reactions , respectively ) . Discrepancies regarding results for each individual model are not only related with differences in the metabolic network but are also dependent on the formulations of the biomass equation ( e . g . the biomass equation in Klebsiella pneumoniae’s model lacks cofactors ) [84] . To validate the predictions of essentiality of metabolic subsystems obtained with GSMs , each experimentally essential gene in DEG was annotated according to its function , using the same system used in GSM’s . This system ( see Methods ) covered more annotations when compared with COG annotations ( 1363 essential metabolic genes annotated in total , when compared with a total of 906 unique metabolic COGs–S1 Fig ) . Moreover , this curated dataset could be directly compared to the modelling results and included some genes annotated in the “General function prediction only” COG category . After annotation , all unique essential genes were identified and the same was done for essential reactions in the models . Both the total number of unique essential reactions ( modelling ) or genes ( experimental ) varies significantly between subsystems ( Fig 3 ) . The total of reactions and genes in each subsystem ( for both modelling and experimental data , respectively ) also varies significantly . To test for independence of the totals of essentials from the sizes of the subsystems , we performed a Fisher’s exact test , and for the majority of subsystems the null hypothesis of dependence was rejected ( p-value less than 0 . 05 ) . In Fig 3 the subsystem of Cofactor and Prosthetic Groups biosynthesis appears isolated with the maximum number of unique essential enzymes both in experimental and modelling data . Three subsystems show a striking difference in the ranking between experimental and modelling data , being more represented in the latter , all related with membrane and cell wall metabolism . Several justifications can be raised for this difference . First and foremost , often in those subsystems the number of different reactions that can be encoded by the same gene is high [85] , ( thus there will be several essential reactions for each essential gene ) . For instance , in the model of Synechocystis there are twelve essential reactions related with fatty acid biosynthesis all encoded by the same gene , fabZ ( sll1605 in the model ) , a gene that is essential in rich medium experimentally , but in that case , counted only once . Similarly , all of the twenty essential fatty acid synthase reactions in the model of M . tuberculosis are encoded by only 3 different genes: Rv1663 and Rv1662 or Rv2940c . There might also be some lack of integration in nomenclature of the reactions in these subsystems in the different GSMs that should not contribute significantly as all models follow the same nomenclature scheme , which was still manually curated for the essential reactions . To further explore essentiality at the level of genes and reactions , the conservation of essentiality across models and experimental datasets was analysed for each reaction and gene in each metabolic subsystem . Strikingly , no reaction was essential in all the models analysed . Three reactions annotated within aromatic amino acids metabolism ( shikimate kinase , 3-phosphoshikimate 1-carboxyvinyltransferase and chorismate synthase ) were essential in all models except for K . pneumoniae . Although there are differences in the models regarding the capacity to uptake aromatic amino acids , this does not justify the observed differences in terms of essentiality . The notable difference between the model of K . pneumoniae and the others is that it lacks cofactors and prosthetic groups in its biomass equation . Those three reactions correspond to the three last steps in the synthesis of chorismate , which is part of the shikimate pathway , which connects central metabolism with aromatic amino acids metabolism . However , this pathway is also the route taken to synthesize several other compounds in the cell , including quinones and folates [86] , which are not present in the biomass equation of Klebsiella , which is likely the cause for the difference in the results as discussed in [84] and above . Reactions essential in several models annotated within the cofactor and prosthetic group biosynthesis subsystem are dihydrofolate synthase , essential in 13 out of the 15 models , and dihydrofolate reductase and NAD kinase , both essential in 12 models . These are related with the biosynthesis of folates and the phosphorylation of NAD to produce NADP . Two reactions involved in the salvage pathways of nucleotides–the biosynthesis of GDP and dTTP—were also essential for 14 models . Three reactions related with the biosynthesis of cell wall components are essential in all models except for B . subtilis and D . ethenogenes—glucosamine-1-phosphate N-acetyltransferase , phosphoglucosamine mutase and UDP-N-acetylglucosamine 1-carboxyvinyltransferase . Acetyl-CoA carboxylase , related with membrane lipid metabolism , is essential in 12 of the 15 models . One reaction not assigned to any subsystem , the HCO3- equilibration reaction , was essential in 11 of all 15 models . To overview the relationship between modelling and experimental results , the conservation of essentiality for each set of results was compared . The inset plot in Fig 4 shows the high correlation obtained between the weighted essentiality for each subsystem between simulated and experimental data . However , there are some differences to be noted . Firstly , regarding the tRNA charging subsystem , it is modelled in only one GSM ( S . oneidensis , Fig 2 ) . This causes this category to appear much more evidently as the second most conserved essential subsystem in experimental data , in contrast with the low result in the simulations of GSMs . It is expected that future GSMs will include this subsystem , but for the sake of comparison of these results , it was excluded from the correlation . Interestingly , regarding the three highly essential reactions in modelling results related with chorismate biosynthesis , these were not found as significantly essential in experimental data . It is known that in minimal media the knock-out of chorismate synthase ( aroC ) in E . coli impairs growth [87] . The non-essentiality of this enzyme in the rich media analysed here indicates that there must be a compound in the media compensating for its absence . It has been shown that , when provided with p-aminobenzoic acid ( PABA ) , para-hydroxybenzoic acid ( PHBA ) or a combination of a precursor from PABA with a non-biological catalyst , the growth of E . coli aroC mutant in M9 minimal medium can be rescued [88] . PABA and its derivatives cannot be uptaken in the genome-scale model of Escherichia coli or any other of the working set . Transporters for these compounds or others that might compensate in rich media for lethal phenotypes in minimal media remain to be integrated in the genome-scale metabolic models and further explored . Again , the subsystem of cofactor and prosthetic groups metabolism has the highest number of reactions appearing as essential in more datasets in experimental data , in accordance with modelling data . Dihydrofolate synthase and reductase ( highest ranking in modelling ) are essential in 11 and 25 experimental datasets , respectively , indicating that either several organisms can overcome the lack of both enzymes by intermediate pathways not yet modelled , or that the experimental assays have produced false negatives ( see the differences between the sizes of the two experimental assays for E . coli in Table 2 discussed above; a special case , dihydrofolate synthase , is further explored in the Discussion ) . NAD kinase appears as highly essential , also in accordance with simulations , in 24 datasets . Several other genes encoding for enzymes essential for the biosynthesis of cofactors are highly essential for cell viability experimentally , even in rich media ( eg . nadE for NAD; coaD and coaE for coenzyme A; hemC for tetrapyrroles; dxr for isoprenoids ) . Cell envelope biosynthesis genes follow as the third most conserved essential functional module , in accordance with the modelling results as well . Based on the premises of evolutionary parsimony and orthology [28] , we proceeded to the analysis at a large scale of the conservation of metabolic genes in the prokaryotic tree of life to infer potential ancestral metabolic functions . Seventy-nine genomes were assayed , representing all the known prokaryotic phyla with a fully sequenced genome ( see Methods for details ) . A phylogenetic tree with these 79 species is available in S2 Fig . All annotated metabolic genes of E . coli K12 were used as queries to search the set of genomes for conserved metabolic genes and respective functions . The results on conservation of metabolic genes are summarized in Fig 5 . The metabolic subsystem with more prevalent genes is Transport , followed by the tRNA charging subsystem with 18 nearly universal aminoacyl-tRNA synthetases . It should be noted though that nearly all of the 41 transport genes conserved in all 79 genomes correspond to ABC transporters ( S1 Table ) with a ubiquitous ATP-binding domain . Two genes involved in oxidative phosphorylation were also conserved in all genomes analysed: atpA and atpD ( ATP synthase subunits alpha and beta , respectively ) . In the subsystem of cofactors and prosthetic group biosynthesis , glutX and sufC were also conserved in all genomes analysed . It should be noted that glutX corresponds to a tRNA charging protein , a glutamyl-tRNA synthetase involved in the biosynthesis of heme , that should have a double annotation; sufC is an atypical cytoplasmic ABC/ATPase , required for the assembly of iron-sulphur clusters [89] . Although the vast majority of genes found conserved in all the genomes analysed correspond to ABC ubiquitous domains , the high conservation ( between 70 and 79 genomes ) of 214 other genes is still prominent . Twenty-eight of those are classified in the subsystem of cofactor and prosthetic group biosynthesis genes ( Fig 5 , S2 Table ) . On a first look , there is no correlation between essentiality ( Fig 4 ) and conservation ( Fig 5 ) at the individual gene level . The same is even more evident in the case of the highly conserved ABC domains in transporters ( S1 Table ) , with the majority ( 20/33 ) not being essential in any dataset in DEG . This substantiates the fact that highly conserved genes are not necessarily highly essential , likely due to genetic redundancy ( see Discussion ) . The only subsystem for which the correlation is positive and significant ( Pearson coefficient 0 . 95 , p-value 1 . 07e-06 ) is Membrane Lipid Metabolism . To assess the unbiased relationship between essentiality and conservation at the individual gene level , we compared data on conservation with the experimental data on essentiality . For this purpose , we calculated a sum score for each gene’s experimental essentiality and compared it with its conservation ( Methods , Fig 6 ) . This allows for measuring the level of evidence of essentiality for each gene individually with precision . In the bottom panel of Fig 6 , the vast majority of genes lie on the bottom area of the plot ( sum of essentiality below zero ) , meaning that they show up in more datasets as non-essential than as essential . However , on the top side of the same plot , where the genes with a positive sum of essentiality lie , the clear majority are shifted to the right–meaning that they are highly conserved . There are some interesting outliers , such as glyS–glycine-tRNA ligase–highlighted in Fig 6 . This corresponds to one instance in which the monophyly rule is violated: E . coli’s type is common for most bacteria , but another type is common to some other bacteria , archaea and eukarya [90 , 91] . Another interesting case is holA , also highlighted , a DNA polymerase delta subunit that is also divergent in its evolutionary history [50 , 92] but which has been poorly studied . Upon splitting the genes into two sets–those with a positive sum of essentiality and those with a null or negative sum , two very different distributions are obtained , shown in the top panel of Fig 6 . A Kolmogorov-Smirnov test states on the independence of both distributions–essential genes are much more likely to be conserved , whereas non-essential genes can or not be highly conserved . The full data on conservation and essentiality sums can be found in S3 File .
The integration done here was , to our knowledge , the first of the kind for a wide variety of phyla of the bacteria and archaea domains , encompassing experimental phenotypic data , results of large-scale computational simulations and sequence data . The experimental genome-scale essentiality data reveal that approximately 25% of prokaryotic essential genes encode for unknown or general functions ( categories S and R in S1 Fig ) , which is a strong warning on the need for experimental studies on the phenotype of these essential proteins for prokaryotic physiology . Moreover , the organisms for which genome-wide essentiality data are available are relatively scarce . While more experimental data are not available , computational models can be valuable tools aiding in the task of decoding prokaryotic metabolism . It is well known that GSMs are limited by the quality of the genome annotations , the formulation of the biomass equation and the pre-defined environmental conditions and other modelling artefacts [27 , 84] , of which the impact in our results we expand below . Here we tried to reduce the impact of these limitations by basing the choice of the models on a large survey of high-quality manually curated models [84] from which 15 balanced , validated , comparable models were chosen , that at the same time represented a wide phylogenetic diversity ( Table 1 ) . The analysis of common patterns of essentiality filtered out the unique essential reactions that might represent specific errors related with individual models . In this manner , we can find core and common features in the overlap of all models . Using GSMs is particularly interesting in this sense , as they allow us to step one level above that of genomes and all their redundancy in the form of isozymes , duplications and confusing phylogenetic events as lateral gene transfers and gene losses . Manually curated GSMs are not mere reflections of genomes–they include thorough revisions of the network and addition of necessary reactions that are encoded by unknown genes or spontaneous chemical transformations . These are also assigned subsystems and counted in our results . Last , as recognized by other authors , the predictive power of comparative analysis can be significantly enhanced by using it within the functional context of pathways and subsystems [19] . The GSMs prediction of which metabolic subsystem has more genes that are commonly essential in multiple species–Cofactor and Prosthetic Group Biosynthesis–was accurate ( Fig 4 ) . The exception of the experimentally highly essential tRNA-charging functionality that was not reflected in the simulations is due to the hindrance of just one model including this subsystem [40] but it should be fixed if all the models represent appropriately this subsystem in the future . However , the analysis done here , by integrating experimental data with several different models still allowed us to identify this subsystem as highly essential and conserved ( Fig 4 and Fig 5 ) . The problem of the unstandardized biomass composition , evidenced by the GSM of K . pneumoniae not predicting any essential reaction involved in cofactor and prosthetic group biosynthesis due to the fact that none of those compounds is present in the biomass equation ( Fig 2 ) is relevant to the results , a subject that we have already addressed in a recent publication [84] . Due to the incompleteness of the networks , it was not possible to complete the biomass equations with the missing cofactors without an impractical manual editing and curation of most models . However , considering the results obtained here , this incompleteness could readily be identified ( Fig 2 ) and did not impair the prediction of an overwhelming majority of essential reactions related with the subsystem of cofactor and prosthetic group biosynthesis ( Fig 4 ) –the overlap of all other models and the experimental data reveals the conserved essentiality of this subsystem . The comparison of the modelling and experimental results can help raise specific hypotheses and directions for more detailed investigation , as discussed above for the essentiality of chorismate synthase in GSMs . In the analysed experimental datasets obtained in rich media , chorismate synthase is not essential , but is has been shown that in minimal media in E . coli the knock-out of its gene impairs growth [87] that can be restored with p-aminobenzoic acid ( PABA ) or derivatives [88] . The transporters for these compounds should be added to the models , and we expect that a curated analysis of experimental studies of auxotrophies in the literature can point several more additions to GSMs that will improve predictions at the gene-level . Moreover , these results point also to the necessity of performing more often essentiality experiments in defined media . On the other direction , genome-scale models can also indicate improvements for experimental assays . At the moment of writing of this manuscript , a new genome-wide screen for Bacillus subtilis was published [93] , where folC ( dihydrofolate synthase ) was found to be essential , confirming our modelling results and contradicting the previous experimental results for B . subtilis that were used here [50] . The analysis of conservation of metabolic genes here was the first performed using a manually curated annotation system for metabolic pathways and subsystems , with the most complete genome-scale metabolic model of a prokaryote to date [49] . Regarding inferences on ancestry , there are some limitations to our approach . We chose to use a single e-value threshold in a local BLAST–this might be a lax threshold , but at the same time it allows us to recover potential very ancient homologs , tracing back all the way to the Last Universal Common Ancestor ( LUCA ) , and not to incur in debates about in and out-paralogs . Moreover , we used only a sample of prokaryotic genomes– 79 –although we made sure to include representatives of all sequenced phyla to date ( S2 Fig ) , and we took a stringent threshold to indicate and not affirm ancestry ( at least 70 out of 79 genomes ) . Looking at the conservation of subsystems ( Fig 5 ) also allows us to overcome the phylogenetic distribution bias . On another note , looking only for universal genes as markers of ancestry can be a limited approach , due to the phenomena of gene loss and lateral gene transfer–ideally , phylogenetic trees should be built for all genes . A recent study used an innovative and large-scale approach to infer on the genome of LUCA , building all trees for protein families based on 1981 prokaryotic genomes [94] . Interestingly , although using a completely alternative approach , the study also concluded on tRNA charging and cofactor metabolism as being ancient subsystems . These findings corroborate that the genes identified here as present in all genomes of all representative phyla are most likely genes present in the last common ancestor [28] . Overall , our results of high conservation of the tRNA charging system , Transport and Oxidative Phosphorylation point to a last common ancestor metabolic network of prokaryotes where most of the nutrients were uptaked with nonspecific transporters at the expense of ATP and in which tRNA charging was already present . The results also suggest that the catalytic role of cofactors and prosthetic groups was a coin highly sought for in early prebiotic systems still maintained today , as this is the most conserved metabolic subsystem after transport and tRNA charging . It is highly likely that genes encoding for enzymes aiding in cofactor biosynthesis were selected for early in primordial evolution , as was suggested elsewhere for the origin of anabolic pathways in prebiotic systems [95] . This work expanded considerably on previous related studies regarding the relationship between gene conservation and essentiality in width and depth . The demonstration that essential genes are more evolutionary conserved that non-essential [22] , corroborated later with more datasets [21] , used the ratio of non-synonymous substitutions to synonymous substitutions in the genomes to estimate conservation ( Ka/Ks ) . Here , 36 experimentally essential datasets were used , that included one Archaea ( Table 2 ) . The conservation was analysed by looking at the presence of each gene in 79 genomes that were manually selected to represent all the phyla with one fully-sequenced genome in the prokaryotic tree of life . Because each gene might be essential in some datasets in DEG , non-essential in others and not assayed in yet others , instead of analysing essential genes separately from non-essential as in the two aforementioned studies , we used a measure of essentiality for each gene ( sum of essentiality ) that takes into account the datasets where it shows up as essential and those where it is non-essential ( Fig 6 ) . The results show that genes with a positive sum of essentiality ( more datasets showing essential than non-essential ) are much scarcer than those with a negative sum; however , it is much more likely that they are highly conserved . For genes with a negative sum of essentiality , there is no tendency for high or low conservation , with a uniform distribution of these genes for all the values of conservation ( corroborating results by Fang et al . [92] ) . We also expanded on previous studies [21 , 22] by integrating in silico simulations and functional assessment of the data , with the conclusion that with the exception of the tRNA charging subsystem , the majority of highly conserved genes related with transport and cofactor biosynthesis are not highly essential ( Fig 6 , S1 and S2 Tables ) . These two subsystems show low single-gene essentiality most likely due to metabolic redundancy caused by known alternative metabolic routes ( for which multiple knock-outs ought to be performed to test for subsystem-level essentiality ) complemented with enzymatic activities not yet known ( supported by the percentage of genes with general function prediction only S1 Fig ) that might also include promiscuous enzymes [96] . The remarkable redundancy of metabolic networks is reflected in the resilience and robustness of prokaryotic life for the billions of years that it has inhabited Earth . | If we tried to list every known chemical reaction within an organism–human , plant or even bacteria–we would get quite a long and confusing read . But when this information is represented in so-called genome-scale metabolic networks , we have the means to access computationally each of those reactions and their interconnections . Some parts of the network have alternatives , while others are unique and therefore can be essential for growth . Here , we simulate growth and compare essential reactions and genes for the simplest type of unicellular species–prokaryotes–to understand which parts of their metabolism are universally essential and potentially ancestral . We show that similar patterns of essential reactions echo phylogenetic relationships ( this makes sense , as the genome provides the building plan for the enzymes that perform those reactions ) . Our computational predictions correlate strongly with experimental essentiality data . Finally , we show that a crucial step of protein synthesis ( tRNA charging ) and the synthesis and transformation of small molecules that enzymes require ( cofactors ) are the most essential and conserved parts of metabolism in prokaryotes . Our results are a step further in understanding the biology and evolution of prokaryotes but can also be relevant in applied studies including metabolic engineering and antibiotic design . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"metabolic",
"networks",
"enzymology",
"genomic",
"databases",
"prosthetic",
"groups",
"network",
"analysis",
"genome",
"analysis",
"enzyme",
"chemistry",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"biochemical",
"cofactors",
"prokaryotic",
"cells",
"metabolic",
"pathways",
"biological",
"databases",
"comparative",
"genomics",
"biochemistry",
"cell",
"biology",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"biosynthesis",
"cellular",
"types",
"genomics",
"metabolism",
"computational",
"biology"
] | 2018 | Metabolic models and gene essentiality data reveal essential and conserved metabolism in prokaryotes |
Among other factors , changes in gene expression on the human evolutionary lineage have been suggested to play an important role in the establishment of human-specific phenotypes . However , the molecular mechanisms underlying these expression changes are largely unknown . Here , we have explored the role of microRNA ( miRNA ) in the regulation of gene expression divergence among adult humans , chimpanzees , and rhesus macaques , in two brain regions: prefrontal cortex and cerebellum . Using a combination of high-throughput sequencing , miRNA microarrays , and Q-PCR , we have shown that up to 11% of the 325 expressed miRNA diverged significantly between humans and chimpanzees and up to 31% between humans and macaques . Measuring mRNA and protein expression in human and chimpanzee brains , we found a significant inverse relationship between the miRNA and the target genes expression divergence , explaining 2%–4% of mRNA and 4%–6% of protein expression differences . Notably , miRNA showing human-specific expression localize in neurons and target genes that are involved in neural functions . Enrichment in neural functions , as well as miRNA–driven regulation on the human evolutionary lineage , was further confirmed by experimental validation of predicted miRNA targets in two neuroblastoma cell lines . Finally , we identified a signature of positive selection in the upstream region of one of the five miRNA with human-specific expression , miR-34c-5p . This suggests that miR-34c-5p expression change took place after the split of the human and the Neanderthal lineages and had adaptive significance . Taken together these results indicate that changes in miRNA expression might have contributed to evolution of human cognitive functions .
Phenotypic differences between species , including human-specific features such as language and tool-making , are thought to have arisen , to a large extent , through changes in gene expression [1] . Indeed , humans and the closest living primate relatives , chimpanzees , display substantial gene expression divergence in all tissues including the brain [2] , [3] . Mechanistically , this divergence might have been caused by mutations in regulatory elements proximal to genes ( cis- effects ) , or changes in expression or sequence of distal regulators ( trans- effects ) . Previous studies focusing on transcription factors ( TFs ) have indicated an excess of human-specific expression divergence for several TFs in the liver [4] and the brain [5] . These findings suggest that changes in TF expression might explain some of human-chimpanzee gene expression divergence . In this study , we investigated the contribution of another type of gene expression regulator , miRNA , to human-specific gene expression divergence . miRNA are short ( 20–23-nucleotide ) , endogenous , single-stranded RNA involved in post-transcriptional gene expression silencing . Mature miRNA function as part of the RNA-induced silencing complex ( RISC ) , mediating post-transcriptional gene expression inhibition [6]–[8] . In animals , the predominant mechanism of miRNA-mediated gene silencing employs complementary base-pairing between the miRNA seed region and the mRNA 3′ UTR region [9] , [10] . This interaction guides RISC to target transcripts , which are consequently degraded , destabilized or translationally inhibited , causing an inverse expression relationship between miRNA and its cognate targets [8]–[12] . miRNA-mediated gene expression silencing has previously been shown to be important for a variety of physiological and pathological processes , ranging from developmental patterning to cancer progression , as well as important neural functions and dysfunctions [7] , [13]–[15] . The roles of miRNA in determining gene expression divergence between species and , in particular , their contribution to expression differences specific to the human brain remains , however , largely unknown .
To assess miRNA expression divergence between human brains and brains of closely related primate species , we measured miRNA levels in two distinct brain regions , the prefrontal cortex ( dorsal-lateral prefrontal region ) and the cerebellum ( lateral cerebellar cortex ) , of humans ( age: 14–58 years ) , chimpanzees ( age: 12–40 years ) and rhesus macaques ( age: 6–15 years ) using high-throughput sequencing ( Illumina ) . In the prefrontal cortex , a brain region known to play a part in the control of high-level cognitive functions , such as abstract thinking and planning [16]–[18] , we measured miRNA expression samples containing RNA pooled from multiple individuals , for each species ( Table S1 ) . To assess technical variation of the sequencing measurements , we prepared and sequenced small RNA libraries twice . In the cerebellum , we sequenced two human samples , one chimpanzee sample and one rhesus macaque sample , all composed from RNA pooled from multiple individuals ( Table S1 ) . We obtained an average of 7 . 6 million sequencing reads per sample , approximately 49% of which could be perfectly mapped to the corresponding reference genome ( Table S2 ) . Based on these data , we detected expression of 413 miRNA covered by least 10 sequence reads in the human prefrontal cortex or cerebellum . To obtain the corresponding miRNA expression estimates for chimpanzees and rhesus macaques , we mapped all annotated human miRNA precursors to the chimpanzee and rhesus macaque genomes , using a combination of reciprocal BLAT , BLAST and liftOver [19]–[21] and extracted mature miRNA sequences using ClustalW2 precursor sequence alignment [22] ( Materials and Methods ) . For 413 miRNA expressed in the human brain , we could unambiguously identify 385 and 390 corresponding genomic locations in the chimpanzee and rhesus macaque genomes , respectively . The vast majority of these miRNA were also detected in chimpanzee ( 375 ) and rhesus macaque ( 366 ) brains ( Table S3 ) . Due to lower quality of the chimpanzee and the rhesus macaque genomes as well as low expression levels of human miRNA with no chimpanzee or macaque orthologs , we omitted these miRNA from further analyses . In all three species , high-throughput sequencing generated highly reproducible miRNA expression measurements , with good positive correlation between technical replicates ( Pearson correlation , r>0 . 99 , p<10−15 ) ( Figure S1 ) . Furthermore , in both brain regions , miRNA expression divergence among species was evidently greater than variation within species ( Figure 1A-1B ) . The extent of miRNA expression divergence followed the phylogenetic relationship among species in both prefrontal cortex and cerebellum , i . e . human and chimpanzee samples clustered as sister species , with macaque samples forming an outgroup . In the human or chimpanzee prefrontal cortex , 325 miRNA were represented by at least 10 sequence reads in at least one technical replicate of one species . All 325 miRNA had orthologs in the chimpanzee genome ( Table S3 ) . Of these , 37 were differently expressed between species in both technical replicates ( Fisher's exact test , p<0 . 01 & fold-change>2 ) . Using an alternative procedure , based on the assumption that sequence read follow a negative binomial distribution , implemented in the edgeR package [23] , 35 miRNA were differently expressed between humans and chimpanzees ( p<0 . 001 & FDR<0 . 01 ) ( Table S4 ) . Thirty one overlapped between the two methods ( binomial test , p<0 . 0001 ) . Using the same criteria , 106 out of 338 miRNA detected in human and rhesus macaque prefrontal cortex were differently expressed between the two species , according to Fisher's test . Eighty-eight out of these 106 miRNA were also classified by edgeR as differently expressed ( Table S4 ) . The vast majority of miRNA expression differences that were found between species in the prefrontal cortex could be reproduced in the cerebellum . Specifically , out of 37 miRNA differently expressed between humans and chimpanzees in the prefrontal cortex , according to Fisher's exact test , 31 ( 84% ) showed consistent expression differences between species in both brain regions ( Figure S2A-S2B ) . Similarly , out of 106 miRNA differently expressed between humans and macaques in the prefrontal cortex , 82 ( 77% ) showed consistent expression differences between the two species in both brain regions ( Figure S2C-S2D ) . In both cases , the agreement between the two brain regions was far greater than expected by chance ( binomial test , p<0 . 0001 ) . Although the prefrontal cortex and the cerebellum are histologically different , previous studies have shown that mRNA expression differences between humans and chimpanzees are largely shared between these two brain regions [24] . Our results indicate that miRNA divergence is similarly shared between the prefrontal cortex and the cerebellum . Furthermore , good agreement of miRNA divergence estimates between the two brain regions supports robustness of our measurements . To further test the robustness of the miRNA divergence estimates obtained using high-throughput sequencing and to overcome potential problems caused by pooling samples from multiple individuals , we measured miRNA expression in the prefrontal cortex of three human , three chimpanzee and two rhesus macaque individuals using miRNA microarrays ( Agilent ) . To exclude possible hybridization artefacts , array probes corresponding to 150 miRNA with sequence differences between humans and chimpanzees and to 313 miRNA with sequence differences between humans and rhesus macaques were masked prior to further analyses . Overall , microarray were less sensitive than sequencing , with a total of 287 miRNA detected as expressed above the default threshold in the human or chimpanzee prefrontal cortex ( Table S5 ) . Concordant with results obtained using high-throughput sequencing , intra-species variation of the microarray miRNA expression measurements was lower than between-species divergence . Further , as in case of the sequencing data , miRNA expression divergence measured by arrays followed the phylogenetic relationship among species ( Figure 1C ) . Further supporting the authenticity of our miRNA divergence estimates , the differences found between humans and chimpanzees or macaques , using high-throughput sequencing , were largely reproduced in microarray experiments . Due to the lower sensitivity of microarray experiments , out of 37 miRNA that were classified as differently expressed between human and chimpanzee prefrontal cortex using sequencing , 12 were detected reliably on the microarrays and 9 showed consistent direction of expression divergence ( Figure 1D ) . Analyzing miRNA expression divergence based on microarray data alone and applying statistical criteria similar to the ones used in sequencing data analysis ( Student's t-test , p<0 . 01 , fold-change>2 ) , four miRNA differed significantly between humans and chimpanzees in the prefrontal cortex . Three of these showed consistent direction of expression change and one passed the significance cutoff level in the sequencing data . Thus , out of 15 miRNA classified as being differently expressed between humans and chimpanzees by at least one methodology , 12 showed consistent direction of expression change ( binomial test , p<0 . 05 ) ( Figure 1D ) . Similarly , out of 106 miRNA differently expressed between human and macaque prefrontal cortex , according to sequencing , 61 were detected by microarrays and 55 showed consistent direction of expression change ( binomial test , p<0 . 01 ) ( Figure 1E ) . Despite overall agreement between the two methodologies , some of the human-chimpanzee miRNA expression differences identified using sequencing were not confirmed by the microarrays . To further test the validity of our results , we measured expression of 6 miRNA in three human and three chimpanzee individuals using a third methodology: quantitative RT-PCR . We chose three types of miRNA differences: ( 1 ) consistent by both methodologies: miR-383 and miR-34c-5p; ( 2 ) significant according to sequencing , but unconfirmed in the microarray experiment: miR-143 and miR-499; ( 3 ) significant according to sequencing , but not detected or masked on the microarrays: miR-184 and miR-299-3p . Quantitative RT-PCR results confirmed expression differences for all miRNA in the first and the third categories , but not for the miRNA in the second category ( Figure S3 ) . Thus , miRNA expression differences that were consistent across methodologies , or large differences that were identified by the sequencing , but masked or undetected on the arrays , are both likely to reflect actual miRNA expression divergence between human and chimpanzee brains . In the prefrontal cortex , we identified 25 miRNA with the human-chimpanzee expression divergence estimates consistent across methodologies or showing large divergence in the sequencing data , but masked or undetected on the arrays ( Table S6 ) . Using rhesus macaque miRNA expression as an outgroup , 13 of the 25 could be assigned to the human evolutionary lineage and 8 to the chimpanzee evolutionary lineage ( Table S6 ) . Requiring significant support by at least two out of three methodologies ( sequencing , microarrays and Q-PCR ) , expression changes in five miRNA ( miR-184 , miR-299-3p , miR-487a , miR-383 and miR-34c-5p ) could be assigned to the human evolutionary lineage and two ( miR-375 and miR-154* ) to the chimpanzee evolutionary lineage ( Figure 2 ) . Six out of 7 miRNA assigned to the human- and the chimpanzee-evolutionary lineages in the prefrontal cortex also showed human- and chimpanzee-specific expression patterns in cerebellum ( Figure 2 ) . Do miRNA expression differences between human and chimpanzee brains contribute to gene expression divergence between these species ? To estimate this , we measured mRNA and protein expression in human and chimpanzee prefrontal cortex: mRNA expression in five individuals of each species using Affymetrix Exon arrays , protein expression in four individuals of each species with two technical replicates using a label-free 2D-MS/MS Thermo-LTQ proteomics methodology ( Table S1 and Table S7 , Materials and Methods ) . Identified miRNA expression differences indeed had a significant negative effect on mRNA and protein expression in the human and chimpanzee prefrontal cortex , i . e . , targets of highly expressed miRNA were down-regulated in the corresponding species ( Figure 3 and Table S8 ) . This effect was significant for differentially expressed miRNA that were identified using both sequencing and microarray methodologies , as well as for miRNA that were identified by sequencing alone ( Figure 3 and Table S8 ) . The significance level of the effect did not depend on the choice of the miRNA target prediction algorithm: In brief , we obtained similar results using TargetScan5 predictions [10] , [25] - based on the presence of conserved miRNA binding sites in mRNA 3′ UTR regions and reported to have good sensitivity and specificity [26] ( Figure 3A-3C and Table S8 ) - as we obtained using PITA ( TOP ) predictions - based on the free energy gained from the formation of the miRNA-target duplex [27] ( Figure 3D-3F and Table S8 ) . Further , the negative effect of miRNA expression differences on mRNA and protein expression could be observed at various miRNA expression level cutoffs . For highly expressed miRNA , the negative effect on their targets' expression levels tended to be more significant ( Table S8 ) . Finally , the negative effect of miRNA on mRNA expression divergence between human and chimpanzee brains could also be reproduced at various mRNA expression divergence cutoffs ( Figure S4E-S4F ) . To assess an overall contribution of miRNA regulation to mRNA and protein expression divergence between human and chimpanzee brains , we calculated the proportion of significant mRNA and protein expression differences that could be negatively associated with miRNA expression differences . Since some of these associations might be caused by factors other than miRNA regulation , we used a number of significant mRNA and protein expression differences showing positive association between miRNA and target genes , as a background . At p<0 . 001 mRNA divergence cutoff ( FDR<2% ) , 68 out of 479 ( 14% ) mRNA , with significant expression differences between human and chimpanzee prefrontal cortex , could be negatively associated with miRNA expression differences . By contrast , 58 ( 12% ) mRNA showed positive association . Thus , 2% of mRNA expression differences between human and chimpanzee brains could be assigned to miRNA regulation . Although this effect appears small , it can be observed consistently at all mRNA expression divergence cutoffs ( Figure S4A-S4B ) . Further , at more stringent mRNA divergence cutoffs , the miRNA regulatory effect became more apparent reaching 4% at p = 0 . 0005 . At the protein level , 26 out of 117 ( 22% ) proteins with significant expression differences between humans and chimpanzees ( FDR<5% ) were negatively associated , and 21 ( 18% ) - positively , with the miRNA expression divergence . Thus , we estimate that 4% of protein expression differences between human and chimpanzee brains could be caused by miRNA . Similarly , the miRNA regulatory effect could be consistently detected at all protein expression divergence cutoffs , and increased to 6% at p = 0 . 001 ( Figure S4C-S4D ) . These estimates are based on the assumptions that negative relationship between miRNA and target gene expression levels in the two species indicates regulation , while a positive relationship does not . Both assumptions might be incorrect . An excess of negative associations between miRNA and their predicted targets might be caused by yet unknown factors , rather than miRNA regulation . On the other hand , positive regulatory relationship between miRNA and target gene expression has been reported [28] , [29] . Further , a positive correlation between miRNA and target gene expression could be caused by indirect regulatory effects [30] . Thus , the actual extent of the effect that miRNA have on gene expression divergence between adult human and chimpanzee brains remains to be estimated . Nevertheless , the consistent and significant negative relationship between miRNA expression , and the expression of their target genes , on both mRNA and protein levels ( Figure 3 and Figure S4 ) , and the consistent excess of negative associations between miRNA and their targets ( Figure S4 ) , demonstrates that miRNA expression divergence does contribute to gene expression divergence between humans and chimpanzees . To assess whether miRNA with expression divergence on the human lineage might be associated with human cognitive functions , we investigated the expression of genes targeted by five miRNA showing human-specific expression , according to multiple methodologies: miR-184 , miR-487a , miR-383 , miR-34c-5p and miR-299-3p ( Figure 2 ) . On the DNA sequence level , these miRNA tend to be conserved: miR-184 mature miRNA sequence is evolutionarily conserved from insects to humans , with only one nucleotide different at 3′end of mature sequence , while miR-383 and miR-34c-5p are classified as broadly conserved and miR-299-3p - as conserved among animal species [25] , [31] . High sequence conservation indicates the functional importance of these miRNA and shows that expression divergence on the human evolutionary lineage is unlikely to be caused by lack of a selection constraint . Notably , genes targeted by these five miRNA were enriched in neural functions . By contrast , no neural-related enrichment was observed for targets of the two miRNA showing chimpanzee-specific expression . Specifically , based on a functional analysis using DAVID [32] , combined targets of the five miRNA with human-specific expression were significantly enriched in terms “signal transduction” , “synaptic transmission” , “cell surface receptor mediated signal transduction” , “neuronal activities” and “cell proliferation and differentiation” ( Bonferroni-corrected p<0 . 05 ) ( Table S9 ) . Targets of miRNA with chimpanzee-specific expression were significantly enriched in terms “nucleoside , nucleotide and nucleic acid metabolism” , “mRNA transcription” and “mRNA transcription regulation” ( Table S9 ) . Similarly , based on the DIANA-mirPath algorithm [33] , targets of miR-184 , miR-487a and miR-299-3p were significantly enriched in KEGG pathways that are related to neural functions ( Table S10 ) . Finding three out of five miRNA with significant target gene enrichment in neural functions was unexpected ( permutation test , p = 0 . 067 ) . Furthermore , miR-184 targets were significantly enriched in “long-term potentiation” pathway – one of the few pathways directly connected to learning and memory formation [34] , [35] . Recent studies have also shown that miR-184 is involved in regulation of neural stem cell proliferation and differentiation [36] . Similarly , targets of miR-299-3p were significantly enriched in the “axon guidance” pathway , which is associated with neuronal cell differentiation and functions . To further test association of miR-184 and miR-299-3p with neuronal functions , we determined their expression patterns in the human and macaque prefrontal cortex by in situ hybridization with specific LNA-probes ( Table S11 ) . Expression of the two miRNA co-localized with expression of NeuN protein , an established vertebrate neuronal-specific marker [37] , [38] ( Figure 4 ) . Thus , even though pathway enrichment results were based on a limited number of miRNAs with species-specific expression in the prefrontal cortex , and although they relied on predicted miRNA-target relationships , both miRNA , with significant target gene enrichment in neuron-related pathways , are indeed preferentially expressed in neurons . Due to the fact that miRNA functions cannot be tested in human or chimpanzee brains , we used two human neuroblastoma cell lines , SH-SY5Y and SK-N-SH , to verify miRNA target predictions and test their functions , as well as their expression changes on the human evolutionary linage . In order to achieve this , we transfected the two cell lines with double-stranded oligonucleotides , which mimic human mature miRNA sequences ( Table S1 , Materials and Methods ) . We tested the effects of all five miRNA showing human-specific expression in brain , as well as effects of the chimpanzee version of miR-299-3p sequence . The effects of each miRNA were assayed 24 hours after transfection using Affymetrix Human Genome U133 Plus 2 . 0 arrays . For each cell line , miRNA regulatory effects were calculated as the difference in expression levels between cells transfected with miRNA analogue and cells transfected with negative control oligonucleotides . For each cell line , transfection with negative control oligonucleotides was carried out in two independent replicates . In both cell lines we observed significant expression inhibition of predicted miRNA targets for all 6 miRNA sequences ( Figure 5 ) . The results were highly consistent between two independent negative control replicates ( Figure S5 and Figure S6 ) and showed significant overlap between the two cell lines ( Figure S7 ) . Further , in both cell lines we observed highly correlated target effects after transfection with human and chimpanzee versions of miR-299-3p . The mature sequence of miR-299-3p contains human-specific C to T substitution at position 10 . While this substitution might affect relative stand selection efficiency during miRNA procession , as well as target selection , we did not find any significant differences in target effects between the human and the chimpanzee versions of miR-299-3p in our experiment ( Figure S8 ) . To capture the majority of possible miRNA targets , we used 9 common target prediction algorithms . In agreement with a previous report [26] , among the 9 algorithms , TargetScan resulted in better agreement between experimental results and target predictions ( Table S12 ) . It is noted , however , that targets predicted by other algorithms showed significant inhibition in transfection experiments . Thus , we used either TargetScan or we combined target predictions and down-regulation in cell line experiments to identify experimentally verified miRNA targets ( Table S13 and Table S14 , Materials and Methods ) . Consistent with results based on computational predictions , the experimentally verified targets of miRNA with human-specific expression showed significant enrichment in certain neuronal functions . Specifically , based on a functional analysis using DAVID [32] , experimentally verified targets of the five miRNA were significantly enriched , compared to genes expressed in the brain and in at least one of the two cell lines ( Fisher's exact test , p<0 . 05 ) , in the following biological processes and KEGG pathways associated with neural functions: “signal transduction” , “synaptic transmission” , “neurotransmitter release” , “adherens junction” and “axon guidance” ( Tables S15 ) . Notably , miRNA-target relationships experimentally verified in the two cell lines was also observed in brain . For each of the five human-specifically expressed miRNA we found inverse correlation between miRNA and target gene expression on the human evolutionary lineage ( Figure 6 ) . This effect was significant for combined targets of the five miRNA , as well as for targets of miR-184 and miR-383 analyzed individually . For the remaining miRNA regulatory effects were not significant , but they did show target expression inhibition . Thus , miRNA-target relationship identified in cell line experiments did allow us to capture miRNA-target relationship , thus explaining some of gene expression changes that took place in the brain on the human evolutionary lineage . While human and chimpanzee evolutionary lineages separated approximately 6–7 million years ago , humans and Neanderthals shared a common ancestor less than half a million years ago [39] . Thus , using Neanderthal data it might be possible to date miRNA expression change more precisely . Although miRNA expression in Neanderthal brain cannot be estimated , signature of positive selection spanning miRNA promoter , or the regulatory region in the human genome , would indicate that expression change might have taken place after human and Neanderthal linage separation [40] . We indeed found a significant excess of human derived SNPs , indicating the presence of positive selection on the human evolution linage after the human-Neanderthal split , in the upstream regions of one out of five miRNA with human-specific gene expression: miR-34c-5p ( Fisher's exact test , Bonferroni corrected p<0 . 05 , Materials and Methods ) . Genome-wide , the possibility of finding a signature of positive selection at this significance level within the upstream region of five randomly chosen miRNA is low ( 1000 permutations , p<0 . 05 ) . Notably , for miR-34c-5p signature of positive selection was located in the putative enhancer region approximately 100kb upstream of the miRNA gene ( Figure 7 ) . Thus , although indirectly , these results indicate that the change in miR-34c-5p with human-specific expression might have taken place after the separation of the human and the Neanderthal evolutionary lineages . Furthermore , positive selection on changes in regulatory regions of this miRNA indicates their potential adaptive significance . Functionally , miR-34c-5p was previously shown to be down-regulated in cancer and Parkinson disease [41]–[44] . We further characterized possible functions of miR-34c-5p in the human brain , based on target genes experimentally verified in cell lines . Compared to the genes expressed in brain , these target genes were significantly enriched , among others , in biological processes “neurotransmitter secretion” and “behaviour” , as well as cellular components “dendrite cytoplasm” , “synapse” and “cell junction” ( Fisher's exact test p<0 . 01 , Tables S16 ) . These findings indicate that changes in miR-34c-5p expression on the human evolutionary linage might have resulted in gene expression changes affecting cognitive functions . In conclusion , despite high sequence conservation of 325 miRNA expressed in the prefrontal cortex , 11% were expressed at significantly different levels in humans and chimpanzees . The vast majority of these differences were also found in cerebellum and were confirmed by microarray and Q-PCR experiments . Importantly , we observed significant inverse relationship between human-chimpanzee miRNA expression divergence and expression divergence of the predicted target genes at both mRNA and protein levels . This indicates that miRNA expression divergence plays an important role in shaping gene expression divergence among species . Approximately half of the miRNA expression differences found in the prefrontal cortex could be assigned to the human evolutionary lineage . These miRNA , as well as their target genes , were conserved at the sequence level . Thus , their expression divergence is unlikely to be explained by a lack of selective constraints . Instead , targets of miRNA with human-specific expression were enriched in neural functions associated with learning and memory pathways , such as “axon guidance” and “long term potentiation” . Potential influence of miRNA divergence on neuronal functions was further confirmed by preferential expression of the corresponding miR-299-3p and miR-184 in cortical neurons , as well as verification of the predicted miRNA-target relationship in two human neuroblastoma cell lines . Based on miRNA-target relationships verified in cell lines , we further demonstrated the effect of miRNA regulation on gene expression changes in brain , on the human evolutionary lineage . Finally , we show that at least one out of five human-specific miRNA expression changes found in brain might have occurred after separation of the human and the Neanderthal evolutionary lineages . Signature of positive selection found in the enhancer region of the miRNA , miR-34c-5p , further indicates that this change might have had adaptive significance . Although these findings do not provide direct evidence that miRNA regulation resulted in human-specific phenotypic adaptations , taken together they indicate that miRNA regulation did contribute to gene expression changes on the human evolutionary lineage and that it affected genes involved in neuronal functions . Further studies are needed to evaluate functional significance of the miRNA-driven transcriptome changes .
Informed consent for the use of human tissues for research was obtained in writing from all donors or their next of kin . All non-human primates used in this study suffered sudden deaths for reasons other than their participation in this study and without any relation to the tissue used . Biomedical Research Ethics Committee of Shanghai Institutes for Biological Sciences completed the review of the use and care of the animals in the research project ( approval ID: ER-SIBS-260802P ) . Human tissue was obtained from the NICHD Brain and Tissue Bank for Developmental Disorders at the University of Maryland , Baltimore , MD . The role of the NICHD Brain and Tissue Bank is to distribute tissue and , therefore , cannot endorse the studies performed or the interpretation of results . All subjects were defined as normal controls by forensic pathologists at the NICHD Brain and Tissue Bank . No subjects who suffered a prolonged agonal state were used . For the prefrontal cortex , samples were taken from the frontal part of the superior frontal gyrus: a cortical region approximately corresponding to Brodmann Area 9 . For all samples , similar proportions of grey and white matter were dissected . Total RNA was isolated from the frozen prefrontal cortex tissue using the Trizol ( Invitrogen , USA ) protocol with no modifications . Prior to low molecular weight RNA isolation , the total RNA from 20 male individuals aged between 14 and 58 years was combined in equal amounts . Low molecular weight RNA was isolated , ligated to the adapters , amplified and sequenced following the Small RNA Preparation Protocol ( Illumina , USA ) with no modifications . Technical replication was completed by independently processing the mixed sample of 20 individuals starting from the low molecular weight RNA isolation step . We carried out the sample preparation and deep sequencing by choosing 5 adult chimpanzee individuals and 5 rhesus macaque individuals following the protocols used for human samples . Details of all samples are given in Table S1 . All original deep sequencing data is deposited in the NCBI GEO database [GSE26545] . Total RNA was isolated using the mirVana miRNA isolation kit ( Ambion ) . 100 ng of each RNA sample were hybridized to Agilent Human microRNA Microarray ( G4471A , Agilent Technologies ) . MicroRNA labelling , hybridization and washing were carried out following Agilent's instructions [45] . Agilent microRNA assays integrate eight individual microarrays on a single glass slide . Each microarray includes approximately 15 k features containing probes sourced from the miRBase public database . The probes are 60-mer oligonucleotides directly synthesized on the array . We used Human miRNA Microarray Version3 , which contains probes for 866 human and 89 human viral microRNAs from the Sanger miRBase v12 . 0 . All samples used in the prefrontal cortex comparison among three species were hybridized to one array . Data from samples hybridized on a single array were processed and analyzed separately to avoid possible batch effects . Images of hybridized microarrays were acquired with a DNA microarray scanner ( Agilent G2565BA ) ; Feature Extraction software v . 10 . 5 . 1 . 1 ( Agilent G4462AA ) was uses for image analysis with default protocols and settings . As miRNA microarray probes are based on human mature miRNA sequences , expression levels of miRNA with sequence differences among species cannot be measured reliably . All probes corresponding to 150 such miRNA between human and chimpanzee and 313 such miRNAs between human and rhesus macaque present on the array were masked prior to expression level analysis , based on the mature sequence comparison . mRNA samples for Affymetrix Human Exon 1 . 0 ST Arrays were prepared following the standard GeneChip Whole Transcript ( WT ) Sense Target Labelling Assay . We processed Exon Array datasets following the steps described in [46] . We processed the human , chimpanzee and rhesus macaque datasets separately . For the human dataset , in order to identify array probes that contain mismatches and multiple locations to human genome ( hg18 ) , we mapped Human Exon 1 . 0 ST probes to the human genome using Bowtie [47] . Based on these alignments , we included probes that matched the genome perfectly and at a single location . For the rhesus macaque and chimpanzee datasets , we applied the same procedure by mapping probes to the rhesus macaque genome ( MMUL1 . 0 ) and the chimpanzee genome ( panTro2 . 1 ) separately . Finally , we chose probes that match the ( i ) human and chimpanzee genomes for human and chimpanzee gene expression comparison and , ( ii ) all three species' genomes for human , chimpanzee and rhesus macaque gene expression comparison . To determine whether the signal intensity of a given probe was above the expected level of background noise , we compared the signal intensity for each probe to a distribution of signal intensities of the anti-genomic probes with the same GC content . Anti-genomic probes are specifically designed by Affymetrix to provide an estimate of the non-specific background hybridization [48] . A probe was classified as detected if its intensity was larger than the 95% percentile of the background probes with the same GC content [48] . To further remove any possible systematic experimental bias among arrays , we performed a PM-GCBG correction and quantile normalization using the R package "preprocessCore" ( http://svitsrv25 . epfl . ch/Rdoc/library/preprocessCore/html/00Index . html ) . Prior to norm-alization , all intensities were log2 transformed . A transcript was classified as detected if more than 80% of probes and at least ten probes per transcript were classified as detected . The intensities of transcripts were summarized by the median polish method . We used the Transcript Cluster Annotations file to map the transcript clusters annotated by Affymetrix to Ensembl genes ( Ensembl54 ) . In cases where multiple transcript clusters mapped to the same gene , we calculated gene expression as the median of all corresponding transcript clusters . None of the transcript clusters overlapped . All original microarray data is deposited in the NCBI GEO database [GSE26545] . Protein sample preparation and 2D LC-MS/MS analysis and peptide identification are described elsewhere [46] . Briefly , proteins were extracted from 100 mg of frozen cerebellar tissue samples . The resulting protein solution was incubatedovernight with Trypsin , followed by ultrafiltration and lyophilization . Lyophilized protein samples werethen dissolved in a loading buffer for the LC-MS/MS analysis . Peptide fractionation and analysis were performed in a pH continuous online gradient ( pCOG ) 2D LC-MS/MS system . Peptide identification was achieved by searching against a database of human peptides ( IPI human v3 . 22 ) and its reversed version representing mock database using SEQUEST program in BioWorks 3 . 2 software suite . A mass tolerance of 3 . 0 Da and one missed cleavage site of trypsin were allowed . Cysteine carboxyamidomethlation was set as static modification and no other modification was checked . All output results were filtered and integrated to proteins by an in-house software “BuildSummary” . Using a false discovery rate ( FDR ) of less than 0 . 5% , all of the matches passing a certain Xcorr and delta CN were regarded as valid . Further , all the peptides that could be assigned to multiple proteins were removed . All identified protein IDs were mapped to Ensembl gene IDs using Biomart . Protein expression of each gene was calculated as a median copy number of all peptides , assigned uniquely to any of the isoforms of the corresponding gene . Genes with more than 5 peptides identified in human and chimpanzee brains were used in the miRNA target effect analysis . Based on this cutoff , we quantified protein expression for a total of 981 genes . The processed protein dataset is provided in Table S7 . For mature miRNA quantification we used the TaqMan MicroRNA Assay ( Applied Biosystems ) system [49] . cDNA was synthesized from 50 ng total RNA from in a 15 µl reaction volume , according to the TaqMan MicroRNA Assay protocol . By using hairpin primers targeting specifically mature miRNAs , reverse transcription was performed using the following program: 30 min at 16°C , 30 min at 42°C , 5 min at 85°C and then held at 4°C . For relative quantification by real time , 1 . 5 µl cDNA were used in a total reaction volume of 20 µl with 1 µl custom TaqMan assay using a Roche LC480 RT PCR System . Each measurement was performed in triplicate for each assay . At least two biological replicates for each species were used . Ct ( threshold cycle ) values of RT PCR were normalized to the endogenous control U6 measured together with the test samples . The relative expression of each miRNA was calculated as log2 of 2-Ct values . We mapped the deep sequencing data following the mapping steps of [50] . For each of the brain sequencing datasets , to remove the adapter sequence at the 3′-end of the sequence reads , all unique sequences were trimmed using the custom trimming procedure . The trimmed sequences of each species were mapped to the corresponding genomes , human genome ( hg18 ) , chimpanzee ( PanTro2 . 1 ) and rhesus macaque ( MMUL1 . 0 ) , using SOAP2 algorithm [51] . Only sequences perfectly matching the genome and with a length ranging from 18 to 28 nucleotides were retained . We quantified the miRNAs expression following the quantification steps of [50] . First , all sequences with at least one read mapping within three nucleotides upstream or downstream of the 5′-position of the mature miRNAs were retained . Then , for each mature miRNA , the sequence with a maximal copy number was designated as the reference sequence . Finally , the expression level of each miRNA was calculated as the sum of the copy number of the reference sequence and the sequences mapping at the same 5′-end position as the reference sequence . Besides the quantification of known miRNAs , novel miRNAs were detected following [50] . Specifically , for the miRNA precursors with one annotated miRNA , small sequences mapping to the opposite arm of the precursor hairpin were analysed . The sequence with the maximal copy number was considered as a novel miRNA candidate . A further criterion required the existence of at least 14 basepairs between an annotated miRNA and a novel miRNA candidate within the precursor hairpin . The quantification process for novel miRNAs was the same as for known miRNAs . Human microRNA information was downloaded from miRBase version 12 [52]-[54] . We used two steps for the ortholog finding; first , we extracted the best precursor orthologs by using a combination of reciprocal BLAT , BLAST and liftOver in chimpanzee and rhesus macaque genomes . Specifically , we mapped all annotated human miRNA precursors to the chimpanzee and rhesus macaque genomes using reciprocal BLAT , BLAST and liftOver , and required one precursor ortholog to be supported by at least 2 out of 3 methods . For reciprocal BLAT , we chose the following parameter configuration: [-stepSize = 5 -repMatch = 2253 -minScore = 0 -minIdentity = 0] . We further required the length of each hit sequence to be more than 70% and less than 130% of the query sequence . For reciprocal BLAST , we chose the parameter configuration [-F F -b 1 –e 10−5] and again required the length of hit sequence to be more than 70% and less than 130% of query sequence . For reciprocal liftOver , we chose the website parameter configuration with Perl LWP module [hglft_minMatch = >0 . 6 , hglft_minSizeT = >0 , hglft_minSizeQ = >0 boolshad . hglft_multiple = >0] and similarly required the length of the hit sequence to be more than 70% and less than 130% of query sequence . We next extracted mature miRNAs based on aligned precursor sequences using ClustalW2 and Muscle , with default parameters . The extracted mature sequence by ClustalW2 and Muscle were highly consistent ( <0 . 1% difference ) . The procedure for identifying differentially expressed miRNAs in deep sequencing data was as follows: We normalized data from two species belonging to the same brain region ( e . g . human and chimpanzee prefrontal cortex ) using quantile normalization . We then used statistical significance , fold-change and detection level as criteria for differential expression ( Fisher's exact test p<0 . 01 , fold-change>2 , at least 10 sequence reads in at least one of the two species ) . We further required that the candidate miRNA should fulfil these criteria in both technical replicates in the prefrontal cortex . Normalization by the number of the total mapped reads ( transcripts per million , TPM ) produced almost identical results [data not shown] . Alternatively , to identify miRNA differentially expressed between humans and chimpanzees or between humans and macaques , we applied a procedure implemented in the edgeR package [23] using the following criteria: p<0 . 001 , FDR<0 . 01 . For identifying differentially expressed miRNA in Agilent miRNA microarray data , a similar approach was used . We first quantile normalized data contained within one Agilent array , and then used both statistical significance and fold-change as criteria for differential expression ( Student t-test , p<0 . 01 , fold-change>2 ) . For the miRNA differently expressed between humans and chimpanzees , we expected targets of miRNA highly expressed in humans to be down-regulated in humans . We first used TargetScan5 [10] , [25] to predict the miRNA targets as this algorithm is reported to have relatively high sensitivity and specificity [26] . To test target effects on the mRNA level , we normalized gene expression between species using quantile normalization and excluded genes with absolute difference between species smaller than 0 . 5 . Using the Wilcoxon signed-rank test , we then compared the expression difference between the targets of miRNA that were highly expressed in humans with targets of miRNA that were highly expressed in chimpanzees . Before applying the Wilcoxon signed-rank test , the genes that were targeted by both miRNA highly expressed in humans and miRNA highly expressed in chimpanzee ( i . e . targets with inconsistent miRNA effects ) were excluded . Due to greater intra-species variation in the protein data , when testing the miRNA target effects on protein expression , we revised the method to use the effect size to represent the expression difference between species . Only genes with absolute effect size greater than one were used in analysis . To check the robustness of detected target effect at both mRNA and protein levels , we used different expression level cutoffs for identification of differentially expressed miRNA , which yielded qualitatively the same result as reported in the main text ( Table S8 ) . We further determined that the target effects could be reproduced using another target prediction algorithm , PITA ( TOP ) [27] . To calculate the percentage of differently expressed genes that could be explained by miRNA expression divergence between human and chimpanzee , we first identified differentially expressed genes with FDR less than 2% at the mRNA level and less than 5% at the protein level ( FDR was estimated using 1000 permutations ) . Among these genes , we determined the percentage that were targeted by differentially expressed miRNA , where at least one miRNA-target gene pair showed expression change in opposite directions . miRNA transfection experiments were conducted on two human derived neuroblastoma cell lines ( SH-SY5Y and SK-N-SH ) ( Table S1 ) . Briefly , cells were plated in 0 . 5 ml of growth medium , without antibiotics , 24 h prior to transfection . miRNA mimics-Lipofectamine 2000 ( Invitrogen ) complexes were prepared freshly before transfection according to the manufacturer's protocol . SH-SY5Y and SK-N-SH cells were transfected in six-well plates using miRNA mimics-Lipofectamine 2000 with a final oligonucleotide concentration of 10 nmol/L . In parallel , negative control transfections with mock oligonucleotides were conducted according to the manufacturer's protocol . For each cell line , transfections with negative control oligonucleotides were carried out in two independent replicates . Cells were harvested after 24 h , total RNA were extracted with Trizol reagent ( Invitrogen ) and further processed and hybridized to Affymetrix Human Genome U133 Plus 2 . 0 arrays following the manufacturer's instructions . The gene expression levels were determined using R RMA package . All original microarray data are deposited in the NCBI GEO database [GSE26545] . We used DIANA-mirPath [33] to determine putative functions of species-specific miRNA . DIANA-mirPath is a web-based computational tool that has been developed to identify molecular pathways potentially altered by the expression of single or multiple microRNAs [33] . The software performs an enrichment analysis of multiple microRNA target genes by comparing each set of microRNA targets to all known KEGG pathways . We chose TargetScan5 and PicTar as target prediction tools and required a score threshold of 6 . 9 ( p<0 . 001 ) ( Table S10 ) . Based on the DIANA-mirPath algorithm , targets of miR-184 , miR-487a and miR-299-3p were significantly enriched in KEGG pathways that are related to neural functions ( Table S10 ) . To test global significance of this result , 1000 simulations were done by randomly choosing five miRNA out of all 325 human miRNA expressed in brain ( Table S3 ) and applying the same test procedure . In 67 out of 1000 simulations , we observed three or more miRNA with enriched KEGG pathways that related to neural functions equal to or larger than the ones observed in the real data ( permutations , p = 0 . 067 ) . In a parallel approach , the DAVID tool for functional annotation of gene sets [32] was used to investigate the putative functions of genes targeted by human-specific or by chimpanzee-specific miRNAs , as predicted by TargetScan . Genes expressed in brain and targeted by human annotated miRNA were taken as background . Significant enriched biological processes based on the PANTHER ( Protein ANalysis THrough Evolutionary Relationships ) Classification System are listed in Table S9 ( Benjamini-Hochberg corrected Fisher's exact test p<0 . 05 ) [55] . Further , we used DAVID to investigate putative functions of experimentally verified target genes of miRNAs with human specific expression , based on our transfection results . Experimentally verified target genes were predicted by TargetScan and were required to show down-regulation by transfection of the corresponding miRNA in at least one of the two cell lines ( Table S13 ) . Experimentally verified target genes expressed in brain were used in functional enrichment analyses . Genes expressed in both brain and at least one of two cell lines were used as a background . Significantly enriched biological processes based on the PANTHER Classification System and KEGG pathways are listed in Table S15 ( Fisher's exact test p<0 . 05 ) . To capture the majority of possible miRNA targets , including non-conserved ones , we combined predictions of 9 algorithms: TargetScan5 [10] , PITA [27] , PicTar [56] , mirSVR [57] , MirTarget2 [58] , microT v3 . 0 [59] , TargetMiner [60] , Antar [61] and 2step-SVM [62] ( Table S12 ) . In order to classify predicted targets as experimentally verified , we calculated target FDR , for each algorithm , based on the inhibitory effect observed in cell line transfection experiments . Specifically , we calculated proportions of predicted target genes and non-target genes inhibited after transfection in both cell lines , at a certain inhibition cutoff ( calculated as the difference in expression between miRNA transfection and the negative control ) . FDR was calculated as the ratio of the proportion of non-target genes passing this inhibition cutoff compared to the total proportion of target genes expressed in the corresponding cell line . The unions of targets predicted by the 9 algorithms at FDR<10% cutoff were used as experimental verified miRNA targets , except for miR-34c-5p , which target FDR was taken at 15% due to a weaker inhibition effect , observed in our transfection experiment ( Table S14 ) . Genes with human-specific and chimpanzee-specific expression were determined by comparison of human-macaque and chimpanzee-macaque expression distances . Genes with greater human-macaque distance were classified to have human-specific expression . Although this requirement is non-conservative , it results in enrichment for genes with human-specific expression . Further , strict identification of human-specific gene expression changes was not a focus of this study . Fisher's exact test was used to determine whether genes with human-specific expression , and showing inverse expression change compared to a given miRNA , were enriched among experimental verified targets of this miRNA . Target genes of this miRNA that were not showing human-specific expression were used as a background . We designed two LNA-probes complementary to miR-184 and miR-299-3p respectively ( Table S11 ) . Hybridizations were performed as described in [63] . Briefly , 10 micrometer-thick tissue sections were collected on Superfrost/plus slides ( Fisher ) . After washing in two changes of excess PBS , sections were acetylated with 0 . 1M triethanolamine/0 . 25% acetic anhydride for 10 minutes and then incubated in humidified bioassay trays for prehybridization at 55°C ( 20–25°C below the Tm of the probe ) for 4 hours in hybridization buffer ( 5xSSC/lx Denhardt's solution/5 mM EDTA/0 . 1% Tween/0 . 1% DHAPS/50% deionized formamide/0 . 1 mg/ml Heparin and 0 . 3 mg/ml yeast tRNA ) [63] , [64] . This procedure was followed by an over-night hybridization step using a DIG-labelled LNA oligonucleotide probe complementary to the target miRNA . Below the temperature of 55°C , sections were rinsed and washed twice in 2xSSC and 3 times in 0 . 2xSSC . The in situ hybridization signal was detected by incubation with alkaline phosphatase ( AP ) -conjugated anti-DIG antibody , using NBT/BCIP as substrate for 3–12 minutes . SNPs used in a genome-wide scan for signals of positive selection in the human lineage , since divergence from the Neanderthal lineage ( Selective Sweep Scan or S SNPs ) [40] , were downloaded from UCSC . Following the published procedure , SNPs were defined as human derived when at least four out of six modern human genomes were derived while all observed Neanderthal alleles were ancestral [40] . An overrepresentation of human derived SNPs in a region would imply that the region had undergone positive selection in the modern human lineage , since divergence from Neanderthals . 50kb sliding windows with a 10 kb step were used to scan the human derived SNPs along the human genome . For five human specific miRNA , we used Fisher's exact test to check overrepresentation of human derived SNPs in each sliding window , with 150 kb region upstream of the annotated miRNA precursor . Four windows in an upstream region of miR-34c-5p were significant at Bonferroni corrected p<0 . 05 . To test the global significance of this result , 1000 simulations were performed by randomly choosing five miRNA precursors out of all 622 annotated human miRNA precursors , and the same test procedure applied . In 44 out of 1000 simulations we observed four or more sliding windows with Fisher's exact test p-values equal to , or smaller than , the ones observed in the real data ( permutation p = 0 . 044 ) . Putative functions of miR-34c-5p targets were determined using CORNA [65] , using experimentally verified target genes of miR-34c-5p as predicted by the 9 aforementioned algorithms , at FDR = 15% . Genes expressed in brain were used as a background . | Humans are remarkably similar to apes and monkeys on the genome sequence level but remain remarkably distinct with respect to cognitive abilities . How could human cognition evolve within such a short evolutionary time ? Among many hypotheses , evolution in expression of a few key regulators affecting hundreds of their target genes was proposed as one possible solution . Here , we tested this notion by studying expression divergence of a specific type of regulatory RNA , microRNA ( miRNA ) , and its effect on gene expression profiles in brains of humans , chimpanzees , and rhesus macaques . Our results indicate that changes in miRNA expression have played a considerable role in the establishment of gene expression divergence between human brains and brains of non-human primates at both mRNA and protein expression levels . Furthermore , we find indications that some of the human-specific gene expression profiles caused by miRNA expression divergence might be associated with evolution of human-specific functions . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"microarrays",
"sequence",
"analysis",
"organismal",
"evolution",
"genomics",
"evolutionary",
"biology",
"human",
"evolution",
"biology",
"computational",
"biology"
] | 2011 | MicroRNA Expression and Regulation in Human, Chimpanzee, and Macaque Brains |
Dengue is a leading cause of morbidity throughout the tropics; however , accurate population-based estimates of mortality rates are not available . We established the Enhanced Fatal Acute Febrile Illness Surveillance System ( EFASS ) to estimate dengue mortality rates in Puerto Rico . Healthcare professionals submitted serum and tissue specimens from patients who died from a dengue-like acute febrile illness , and death certificates were reviewed to identify additional cases . Specimens were tested for markers of dengue virus ( DENV ) infection by molecular , immunologic , and immunohistochemical methods , and were also tested for West Nile virus , Leptospira spp . , and other pathogens based on histopathologic findings . Medical records were reviewed and clinical data abstracted . A total of 311 deaths were identified , of which 58 ( 19% ) were DENV laboratory-positive . Dengue mortality rates were 1 . 05 per 100 , 000 population in 2010 , 0 . 16 in 2011 and 0 . 36 in 2012 . Dengue mortality was highest among adults 19–64 years and seniors ≥65 years ( 1 . 17 and 1 . 66 deaths per 100 , 000 , respectively ) . Other pathogens identified included 34 Leptospira spp . cases and one case of Burkholderia pseudomallei and Neisseria meningitidis . EFASS showed that dengue mortality rates among adults were higher than reported for influenza , and identified a leptospirosis outbreak and index cases of melioidosis and meningitis .
Dengue is a major public health problem worldwide . While most dengue virus ( DENV ) infections are asymptomatic or result in a mild acute febrile illness ( AFI ) , some are life-threatening due to plasma leakage [1 , 2] . With no antivirals to treat dengue or prevent its severe manifestations [3] , early recognition of severe dengue and timely supportive care is used to reduce mortality [4–6] . Several dengue vaccines are in late stage clinical trials and one was recently licensed in several countries [7] . Decisions regarding their use will depend on vaccine performance and safety , and reduction of disease burden , including deaths . Globally , an estimated 3 . 9 billion people are at risk of DENV infection , and 96 million dengue cases are estimated to have occurred in 2010 alone [8] . Despite its global presense , robust estimates of population-based dengue mortality rates are lacking . Most estimates have been derived from passive surveillance data [9–14] or hospital-based , retrospective case reviews [15 , 16] . During epidemic periods , these methods have produced annual mortality rates that ranged from 0 . 30–0 . 59 deaths per 100 , 000 population . However , these approaches have not been validated as to under recognition due to misdiagnosis or under reporting [14] . Dengue has been endemic in Puerto Rico since the late 1960s [17 , 18] , and after the first deaths were reported in 1986 , surveillance for deaths due to dengue was established in 1987 [19] . Mortality data have been collected through the island-wide Passive Dengue Surveillance System ( PDSS ) , a hospital-based Infection Control Nurse Dengue Surveillance System ( ICNDSS ) that operated until 2007 , and review of death certificates . Evaluations of these systems identified misdiagnosis and underreporting of cases , and failure to include dengue on death certificates of known laboratory-positive dengue cases [14 , 20] . Few suspected fatal cases had tissue specimens or appropriately timed pre-mortem serum specimens for diagnostic testing , which resulted in a high proportion of indeterminate diagnostic results [14 , 19–21] . In 2009 , the Centers for Disease Control and Prevention Dengue Branch ( CDC-DB ) , Puerto Rico Department of Health ( PRDH ) , Instituto de Ciencias Forenses de Puerto Rico ( in English , Puerto Rico Institute of Forensic Sciences [PRIFS] ) , Demographic Registry of Puerto Rico , and CDC Infectious Diseases Pathology Branch ( CDC-IDPB ) established the Enhanced Fatal Acute Febrile Illness Surveillance System ( EFASS ) to define mortality due to dengue-like AFI , and determine the etiology of these cases . We describe the findings from the first three years of EFASS .
This project underwent CDC institutional review and formal institutional review board review was not required since the case-patients were deceased . Because cases were reported in the context of public health surveillance , the informed consent of patients’ families was not sought . Patient identifiers were removed from the dataset prior to analysis . EFASS used enhanced surveillance to detect dengue-like AFI deaths , improve reporting , and standardize collection of specimens at autopsy . While PDSS provided retrospective diagnostic data on fatal suspected dengue cases reported early in their illness , the primary source of EFASS cases was reporting by participating epidemiologists , pathologists , and registry statisticians . Specifically , they were asked to report and submit samples from all fatal cases whose death occurred during or immediately following a dengue-like AFI defined by the presence of fever ( body temperature ≥38 . 0C axillary ) or history of fever for ≤7 days . This included deaths with pre-defined diagnostic codes on the medical record , autopsy report or death certificate ( S1 Appendix ) ; the list of ICD codes was developed in 2009 after a review of the 1994–2007 fatal laboratory-positive dengue cases was conducted . Surveillance was enhanced by collaboration with pathologists and epidemiologists at hospitals most likely to encounter severe dengue cases , and included training and provision of standardized protocols . Dengue-like AFI fatalities that occurred at home or within 24 hours of hospital admission and referred to PRIFS were included . Collaborators were contacted weekly , and death certificates , PDSS , and National Notifiable Diseases Surveillance System ( NNDSS ) were routinely queried to identify suspected cases . Once a suspected case was identified , serum , whole blood , and tissue specimens were obtained , and PRIFS pathologists completed a Surgical Pathology and Autopsy Report ( SPAR ) ( S2 Appendix ) . Cases with history of respiratory failure had a nasopharyngeal swab submitted for testing . Data was abstracted from medical records of all health care visits during the illness for laboratory-positive dengue cases using a standard instrument that captured demographic characteristics , past medical history , clinical course , and management . Serum specimens were tested by a DENV-serotype specific real time , reverse transcriptase-polymerase chain reaction ( rRT-PCR ) assay [22] , an anti-DENV IgM enzyme-linked immunosorbent assay ( MAC ELISA ) [23] , and an anti-DENV IgG ELISA [24] . Serum specimens were also tested by anti-West Nile virus ( WNV ) MAC-ELISA and , if positive , WNV-specific rRT-PCR and 90% plaque reduction neutralization tests ( PRNT90 ) were performed [25] . Serum specimens with sufficient volume were sent to CDC Bacterial Special Pathogens Branch and tested for Leptospira IgM using the ELISA ImmunoDOT kit ( GenBio , Inc . , San Diego , CA ) . Acute specimens were tested for nucleic acid by rRT-PCR and 20 Leptospira reference antigens representing 17 serogroups by microscopic agglutination test ( MAT ) [26] . RNA was extracted from nasopharyngeal specimens and tested for Influenza A and B viral genome by rRT-PCR [27] . Tissue specimens were tested at the CDC-IDPB for DENV antigen or nucleic acid by immunohistochemistry ( IHC ) and RT-PCR , respectively [28] . If clinical presentation or histopathology were suggestive of another etiology pathogen-specific diagnostic testing was performed [26 , 29] . A fatal suspected dengue-like AFI case had a dengue-like AFI that immediately preceded death in a Puerto Rico resident . A fatal laboratory-positive dengue case was a suspected case with DENV nucleic acid in serum or tissue; DENV antigen in tissue; IgM seroconversion in paired specimens; or IgM in a single specimen . A fatal laboratory-negative dengue case was a suspected case with no molecular , immunodiagnostic or IHC markers of DENV infection . A fatal laboratory-indeterminate dengue case was a suspected case with no DENV nucleic acid or anti-DENV IgM in the acute serum specimen ( collected ≤5 days post-illness onset [DPO] ) and no available convalescent serum specimen ( ≥6 DPO ) . A fatal dengue co-infection was a fatal suspected dengue-like AFI case with DENV nucleic acid in serum or tissue and another pathogen detected by PCR or IHC . A primary DENV infection was a fatal laboratory-positive dengue case without anti-DENV IgG in the acute serum specimen [30] and a secondary DENV infection had anti-DENV IgG in the acute specimen . A fatal laboratory-confirmed leptospirosis case was a suspected dengue-like AFI case with ≥4-fold increase in MAT titers in paired specimens , MAT titer ≥800 in a single specimen , or detection of Leptospira spp . nucleic acid in serum by PCR or antigen in tissue by IHC . A probable fatal leptospirosis case had a MAT titer >200 but <800 in a serum specimen . Dengue fever ( DF ) , dengue hemorrhagic fever ( DHF ) and dengue shock syndrome ( DSS ) were defined according to the 1997 World Health Organization ( WHO ) guidelines [31] ( Table 1 ) . Dengue , dengue with warning signs , and severe dengue were defined according to 2009 WHO guidelines [2] . Definitions for severe dengue , other clinical features and medical complications are shown in Table 1 . Frequencies were calculated for demographic , clinical and laboratory features of fatal laboratory-positive dengue cases . Rates of fatal laboratory-positive dengue cases per 100 , 000 Puerto Rico population were calculated by age group , sex , and municipality using US Census data [45] . Incidence rate ratios ( IRR ) were calculated to compare females to males . Case fatality rates ( CFR ) were calculated by dividing the number of fatal laboratory-positive dengue cases by PDSS laboratory-positive dengue cases . Statistical differences in proportions were tested by Chi-square or Fisher's exact tests . Differences between municipality-specific fatal laboratory-positive dengue cases and PDSS laboratory-positive dengue cases were examined by calculation of Pearson correlation coefficients . A geographically weighted regression model was used to determine if the number of fatal cases differed from the expected based on PDSS laboratory-positive dengue cases in the municipality and neighboring municipalities . Data analyses were conducted using STATA ( Stata Corporation , College Station , TX ) and ArcGIS ( Environmental Systems Research Institute , Redlands , CA ) ; maps were created using ArcMap .
During 2010–2012 , a total of 311 fatalities following a dengue-like AFI were detected by EFASS and 40 , 881 suspected dengue cases were reported to PDSS , of which 17 , 929 ( 44% ) were dengue laboratory-positive . Of all fatalities detected , 146 ( 47% ) were identified and reported by PRIFS pathologists , 93 ( 30% ) by death certificate review , 50 ( 16% ) by hospital epidemiologists , 15 ( 5% ) by PDSS , four ( 1% ) by NNDSS , and three ( 1% ) by chart review as part of another study . Serum and tissue specimens were available for 148 ( 48% ) cases , serum alone for 138 ( 44% ) cases , tissue alone for 16 ( 5% ) cases , and 9 ( 3% ) cases had no diagnostic specimens . Of the 164 cases ( 53% ) with tissue , one case was not tested because of sample quality . Of evaluable cases , 159 ( 98% ) had liver , 156 ( 96% ) lung , 155 ( 95% ) kidney , 142 ( 87% ) spleen , 98 ( 60% ) lymph nodes , and 66 ( 41% ) intestine . A nasopharyngeal swab was submitted for 27 ( 9% ) cases . A pathogen was identified in 120 ( 40% ) of 302 cases with a diagnostic specimen . A pathogen was more likely to be identified in cases with tissue specimens than in those without ( 69% versus 45% respectively , P <0 . 0001 ) . Overall , 58 ( 19% ) fatal cases were dengue laboratory-positive , 167 ( 54% ) were dengue laboratory-negative , and 77 ( 25% ) were dengue laboratory-indeterminate . Other pathogens identified included: Leptospira spp . in 34 ( 11% ) cases ( 32 confirmed , two probable ) ; Staphylococcus spp . in nine ( 3% ) cases; Streptococcus spp . in nine ( 3% ) cases; influenza A virus in three ( 1% ) cases; and one case each with Neisseria meningitidis , Burkholderia pseudomallei , Proteus spp . , Clostridium perfringens , Cryptococcus neoformans , Klebsiella pneumoniae , and an unidentified Gram-positive coccus . Of the 58 fatal laboratory-positive dengue cases , 53 ( 91% ) were DENV RT-PCR positive in tissue , serum or both; four ( 7% ) were anti-DENV IgM positive in a single serum specimen; and one ( 2% ) demonstrated anti-DENV IgM seroconversion in paired specimens ( Table 2 ) . Autopsies were performed on 26 ( 45% ) of the 58 fatal laboratory-positive dengue cases and DENV was most commonly identified by IHC or RT-PCR in liver ( 18/26 , 69% ) , lung ( 15/22 , 68% ) , and kidney ( 9/24 , 38% ) tissue . Of 43 cases with rRT-PCR positive serum specimen , 26 ( 60% ) were DENV-1 , 16 ( 37% ) DENV-4 , and one ( 2% ) DENV-2; similar to DENV-type distribution in PDSS during the same time period [18] . Among the 36 fatal laboratory-positive dengue cases with an available acute serum specimen , 30 ( 83% ) had a secondary DENV infection and 6 ( 17% ) had primary infection . Five of the DENV RT-PCR positive cases had co-infection with another pathogen , including Leptospira species ( 4 cases ) [46] and Streptococcus species ( 1 case ) . Fatal laboratory-positive dengue cases occurred in the months with increased PDSS dengue reporting ( Fig 1 ) . The dengue mortality rate was 1 . 05 per 100 , 000 population in 2010 , 0 . 16 in 2011 and 0 . 36 in 2012; the 3-year average mortality rate was 0 . 52 dengue deaths per 100 , 000 . The overall CFR for the three-year period was 0 . 32% and varied from 0 . 38% during the 2010 epidemic [18] , ; to 0 . 39% in 2011 , a non-epidemic year , to 0 . 22% in 2012 , an epidemic year . The median age of fatal laboratory-positive dengue case-patients was 46 years ( range: 5 months–89 years ) . Six ( 10% ) case-patients were <20 years old , 39 ( 67% ) 20–64 years old , and 13 ( 22% ) ≥65 years old . EFASS case-patients were significantly older than laboratory-positive dengue case-patients reported to PDSS ( median 46 vs . 18 years , P < 0 . 001 ) and a higher proportion of fatal laboratory-positive dengue case-patients were adults ( 90% vs . 49% , respectively , P < 0 . 001 ) . The majority of fatal laboratory-positive dengue case-patients were female ( Table 3 ) , and were significantly more likely to be female than laboratory-positive dengue cases reported to PDSS during the same period ( 59% vs . 45% , P <0 . 05 ) . In 2010 , rates of fatal laboratory-positive dengue were 1 . 3 times higher among females than males; 1 . 19 and 0 . 90 cases per 100 , 000 population , respectively ( female-to-male IRR = 1 . 3; 95% confidence interval [CI] = 0 . 70–2 . 50 , P = 0 . 20 ) . Slightly less than half ( 27 , 47% ) of fatal laboratory-positive dengue case-patients were obese , similar to that reported in Puerto Rico [47] . Most fatal laboratory-positive dengue case-patients had more than one pre-existing medical condition ( 35 , 60% ) . The most common included: diabetes mellitus ( 23 , 40% ) , asthma ( 13 , 22% ) , cardiovascular disease ( 9 , 16% ) , thyroid disease ( 8 , 14% ) , psychiatric disease ( 8 , 14% ) , and rheumatologic conditions ( 8 , 14% ) . Fatal laboratory-positive dengue case-patients were residents of 35 of the 78 Puerto Rico municipalities , and most ( 53 , 91% ) were born in Puerto Rico ( Table 3 ) . Fatal laboratory-positive dengue rates were highest in 2010 in Maunabo ( 0 . 82 per 10 , 000 residents ) , Maricao in 2011 ( 1 . 59 ) , and Adjuntas in 2012 ( 0 . 51 ) . In all years , there was a positive correlation between the number of PDSS laboratory-positive dengue cases and the number of fatal laboratory-positive dengue cases in a municipality ( R = 0 . 56 , 0 . 59 , 0 . 62 , and 0 . 73 in 2010 , 2011 , 2012 , and overall , respectively ) ( Fig 2 ) . The number of fatal laboratory-positive dengue cases detected was no more than expected based on municipality-specific , laboratory-positive dengue incidence rates . However , there were fewer than expected fatal laboratory-positive dengue cases detected in 11 of 78 municipalities in individual years . Only Quebradillas and Patillas had fewer than expected fatalities in all years . All 58 fatal laboratory-positive dengue case-patients sought care at least once during their illness; 25 ( 43% ) had two healthcare visits , and seven ( 12% ) had three or more visits . The median interval between fever onset and arrival at the first visit was 3 . 5 days ( range: 0–8 . 5 days ) . Of the 49 ( 84% ) cases with a medical record for review , at the first visit , most ( 71% ) had dengue fever , while some had severe dengue ( 35% ) or DHF ( 18% ) ( Table 4 ) ; the majority ( 67% ) had warning signs for severe dengue , the most common being persistent vomiting ( 21 of 33 , 64% ) , abdominal pain ( 15 , 45% ) , and mucosal bleeding ( 14 , 42% ) . The leading diagnoses during the first visit were dengue , viral syndrome , gastroenteritis , and urinary tract infection . At the first visit , 17 ( 35% ) were admitted to the hospital , 7 ( 14% ) were transferred to another hospital , 7 ( 14% ) died in the ED , and 18 ( 37% ) were discharged home . Of those transferred to another hospital , six ( 86% ) were admitted and died in the ICU , and one died in the ED . Among the 18 case-patients discharged home after their first visit , seven ( 39% ) had at least one recorded warning sign , including compensated shock or hemorrhagic manifestations . Median age was 36 . 6 years ( range 0 . 6–74 . 9 ) , 14 ( 78% ) were female , and 16 ( 89% ) had at least one chronic medical condition . Two died at home after being discharged; diagnoses of dengue and moderate dehydration , and acute gastroenteritis . The remaining 16 returned to a hospital on average 2 days ( range 0 . 0–3 . 5 days ) after discharge . Nine ( 56% ) died during their second visit , four ( 25% ) were transferred to another facility , and three ( 19% ) were again discharged home . Overall , the median interval between illness onset and death was 7 . 1 days ( range: 1 . 2–28 . 4 days ) ; in six ( 10% ) cases the interval exceeded 14 days . Most ( 43 , 74% ) case-patients died as an inpatient in a hospital; however , 12 ( 21% ) died in the Emergency Department prior to hospital admission , and three ( 5% ) died at home: one after a 2-day hospitalization , and two after being seen in an ED ( Table 4 ) . Of the 43 case-patients who died after being admitted , 33 ( 57% ) died in the intensive care unit , and 10 ( 17% ) died in an inpatient ward . The median interval from hospital admission to death was 1 . 9 days ( range: 0 . 1–28 . 8 days ) . Only 25 of 58 ( 43% ) fatal laboratory-positive dengue case-patients had dengue , DHF , DSS , or dengue-like syndrome listed as primary ( 17 cases ) or contributing ( 8 cases ) cause of death on their death certificate . The five most common primary causes of death included dengue ( 29% ) , viral syndrome ( 16% ) , cardiorespiratory failure ( 12% ) , sepsis ( 9% ) , and thrombocytopenia ( 5% ) . Most of the 55 laboratory-positive dengue case-patients who died in hospital had signs and symptoms consistent with dengue ( 86% ) or severe dengue ( 82% ) , but only 62% had that clinical diagnosis ( Table 4 ) . Of those who met criteria for severe dengue by the time of death , 38 of 45 ( 84% ) had severe plasma leakage , 25 ( 56% ) had severe bleeding , and 22 ( 49% ) had severe organ impairment . Of note , hemoconcentration was documented in only a few case-patients even when hematocrits were performed over time . However , 34 ( 62% ) case-patients who died in hospital had an effusion and most ( 62% ) case-patients with effusions had acute respiratory failure or ARDS ( Table 5 ) . The majority of case-patients who died in a hospital ( 86% ) were bleeding , and 25 ( 46% ) had severe bleeding: 19 ( 35% ) gastrointestinal , 14 ( 26% ) pulmonary , six ( 11% ) vaginal , and six ( 11% ) intracranial ( Table 5 ) . Of those with severe bleeding , 14 ( 56% ) received a platelet transfusion while 11 ( 44% ) received a blood transfusion; blood transfusion was more likely to occur with increased hospital stay ( median 5 . 2 vs . 1 . 3 days , respectively; P <0 . 001 ) . Of the 26 case-patients who received a platelet transfusion , three ( 12% ) had no recorded bleeding . The best predictor of platelet transfusion was low platelet count [median 13 , 000 ( range: 7 , 000–37 , 000 ) vs . 55 , 000 ( range: 8 , 000–224 , 000 ) cells/mm3 , recipients versus non-recipients , respectively; P <0 . 001] . Other severe clinical outcomes among the 55 laboratory-positive dengue case-patients who died in hospital included acute hepatitis ( 76% ) , acute renal failure ( 18% ) , cholecystitis ( 9% ) , acute liver failure ( 9% ) , myocarditis ( 4% ) , metabolic acidosis ( 66% ) , prolonged shock ( 53% ) , and acute respiratory failure ( 55% ) ( Table 5 ) . All received intravenous crystalloids , 40% received a diuretic , and 29% received intravenous colloids . Twenty-five ( 45%; all adults ) received intravenous corticosteroids . Those given steroids had a lower median platelet count than those not given steroids ( 16 , 000 vs . 47 , 000 cells/mm3; respectively , P <0 . 01 ) , and were three times more likely to have a hospital acquired infection ( 75% vs . 25% , P <0 . 05 ) .
Enhanced surveillance for dengue deaths showed the majority were not reported to the standard dengue surveillance system and most did not have “dengue” coded on the death certificate . Identification of these unrecognized deaths resulted in a 2 to 3-fold higher dengue mortality rate than previously reported [14 , 17 , 20 , 21 , 48 , 49] . EFASS demonstrated the importance of appropriate diagnostic testing of tissue and serum to make the correct diagnosis in deaths from a dengue-like acute febrile illness . In addition , EFASS showed its ability to identify unrecognized deaths from other pathogens of public health importance . The EFASS estimated age-specific annual dengue mortality rates were comparable to those from other infectious diseases in the US , including influenza [50 , 51] . However , in contrast to influenza , most dengue deaths occurred among adults 19–64 years of age . The estimated average annual influenza-associated US death rate is 2 . 4 per 100 , 000 residents ( range: 0 . 4–5 . 1 ) . In most years , 88% of these deaths are among persons aged ≥65 years [51 , 52]; 17 . 0 deaths per 100 , 000 ( range: 2 . 4–36 . 7 ) . Influenza death rates among persons <19 years and 19–64 years are 0 . 1 ( range: 0 . 1–0 . 3 ) and 0 . 4 ( range: 0 . 1–0 . 8 ) per 100 , 000 , respectively . In comparison , EFASS estimated that dengue mortality in 2010 was 0 . 42 , 1 . 17 , and 1 . 66 per 100 , 000 persons aged <19 years , 19–64 years and ≥65 years , respectively . Most fatal laboratory-positive dengue case-patients appeared to have timely access to healthcare . However , many ( ~40% ) were sent home after their first ED visit with warning signs of severe dengue . Although the majority sought care again within 48 hours , two died at home . Most case-patients who died in a hospital had severe plasma leakage , severe bleeding , or both , and most received inotropes and half received a platelet transfusion . Although bleeding was present in the majority who received platelets , half of those with severe bleeding did not receive red blood cells . A large proportion of case-patients received corticosteroids , which are not considered of benefit in dengue [2 , 53] . As reported by others , we found an increased risk of hospital-acquired infections in these patients [54 , 55] . Dengue deaths often occur among patients with comorbidities [14 , 19] . Nearly half of case-patients were obese and over half had more than one chronic medical condition; prevalences similar to those found in the Puerto Rican adult population [47 , 56] , with the exception of diabetes and asthma . The prevalence of diabetes in case-patients was nearly four times that of the adult population , and asthma was twice as prevalent . Adult diabetics have been over-represented in other fatal case series [19 , 57] , and a recent meta-analysis found diabetes was associated with increased risk of severe dengue [58] . As many endemic areas have reported a substantial proportion of dengue cases in adults , healthcare providers should be attentive to dengue patients with these comorbidities [2] . Some patients developed acute liver or renal failure or had atypical presentations [19 , 57 , 59 , 60] . Acute renal failure ( ARF ) affected ~20% of case-patients though none had pre-existing renal disease and 80% were non-elderly ( median age 49 years ) . However , dengue patients with severe dengue , diabetes or secondary infections are known to be at risk for developing acute kidney injury [61] . Six of the ten ARF cases had at least one risk factor and two were co-infected with Leptospira spp . One of the four ARF case-patients without risk factors was an infant with abdominal compartment syndrome and multiple organ dysfunction . While more sensitive than PDSS , EFASS may not have detected all fatal laboratory-positive dengue cases . For example , a few rural municipalities had fewer deaths than expected . In the case of Patillas , this may have been due to higher dengue case-reporting to PDSS because of an enhanced dengue surveillance project conducted prior to EFASS [62] . Alternatively , individuals in rural municipalities who died at home and were not known to have an AFI would not have been identified . These factors may have led to lower case ascertainment and estimated dengue mortality . Although we increased the proportion of suspect cases with an etiologic diagnosis by obtaining tissue and convalescent serum specimens , about one quarter of cases were dengue laboratory-indeterminate , and were not counted as fatal laboratory-positive dengue cases even if dengue was listed on their death certificate . Hence , our final dengue mortality estimate should be considered conservative . EFASS demonstrated the feasibility and importance of enhanced surveillance for dengue deaths , and found a previously unrecognized high dengue mortality in Puerto Rico that was higher than rates observed in other dengue endemic regions during this time period [9–14] . Establishment of EFASS-like systems in selected dengue endemic countries would go a long way towards obtaining robust estimates of the global burden of deaths due to dengue , and identify areas for improvement in clinical care of patients with severe dengue . | Dengue is a major public health problem in the tropics . Despite its global importance , population-based mortality rates attributable to dengue are largely unknown . Dengue vaccines are now in late stage clinical trials and one vaccine has been licensed in several countries . Evidence-based decisions regarding the future use of dengue vaccines will depend on robust estimates of disease burden which should include mortality . To estimate mortality due to dengue in Puerto Rico , where dengue is endemic , we developed an enhanced surveillance system to detect fatalities due to a preceeding dengue-like acute febrile illness using more sensitive case identification and laboratory methods than the previous passive method . This surveillance system found the dengue mortality rate was 1 . 05 per 100 , 000 Puerto Rico residents in 2010 , the highest rate ever detected . Among adults aged 19–64 years , mortality from dengue ( 1 . 17 deaths per 100 , 000 ) was higher than from other infectious diseases , including influenza . The utility of this enhanced surveillance system was further proven through the identification of an outbreak of leptospirosis as well as detection of other diseases of public health importance , including melioidosis and meningitis . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"death",
"rates",
"reverse",
"transcriptase-polymerase",
"chain",
"reaction",
"medicine",
"and",
"health",
"sciences",
"clinical",
"laboratory",
"sciences",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"demography",
"pathogens",
"tropical",
"diseases",
"geographical",
"locations",
"microbiology",
"north",
"america",
"signs",
"and",
"symptoms",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"population",
"biology",
"nosocomial",
"infections",
"caribbean",
"transfusion",
"medicine",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"artificial",
"gene",
"amplification",
"and",
"extension",
"medical",
"microbiology",
"dengue",
"fever",
"microbial",
"pathogens",
"molecular",
"biology",
"hematology",
"people",
"and",
"places",
"puerto",
"rico",
"population",
"metrics",
"diagnostic",
"medicine",
"polymerase",
"chain",
"reaction",
"hemorrhage",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"vascular",
"medicine",
"blood",
"transfusion",
"organisms"
] | 2016 | Enhanced Surveillance for Fatal Dengue-Like Acute Febrile Illness in Puerto Rico, 2010-2012 |
Helminth infection is common in malaria endemic areas , and an interaction between the two would be of considerable public health importance . Animal models suggest that helminth infections may increase susceptibility to malaria , but epidemiological data has been limited and contradictory . In a vaccine trial , we studied 387 one- to six-year-old children for the effect of helminth infections on febrile Plasmodium falciparum malaria episodes . Gastrointestinal helminth infection and eosinophilia were prevalent ( 25% and 50% respectively ) , but did not influence susceptibility to malaria . Hazard ratios were 1 for gastrointestinal helminth infection ( 95% CI 0 . 6–1 . 6 ) and 0 . 85 and 0 . 85 for mild and marked eosinophilia , respectively ( 95% CI 0 . 56–1 . 76 and 0 . 69–1 . 96 ) . Incident rate ratios for multiple episodes were 0 . 83 for gastro-intestinal helminth infection ( 95% CI 0 . 5–1 . 33 ) and 0 . 86 and 0 . 98 for mild and marked eosinophilia ( 95% CI 0 . 5–1 . 4 and 0 . 6–1 . 5 ) . There was no evidence that infection with gastrointestinal helminths or urinary schistosomiasis increased susceptibility to Plasmodium falciparum malaria in this study . Larger studies including populations with a greater prevalence of helminth infection should be undertaken .
Helminth and malaria endemic areas frequently coincide , and an interaction between the two would be of considerable public health importance [1] , [2] . Animal models suggest that helminth infections may alter susceptibility to malaria , although the results are conflicting [3] , [4] . In studies conducted among malaria endemic populations in Africa , helminths have been reported to increase susceptibility to clinical malaria [5] , [6] , reduce the risk [7] or make no difference [8] . Studies in Thailand suggest helminths might increase the risk of non-severe malaria , but reduce the risk of cerebral malaria [9] , [10] . The Th2 cytokine milieu induced by helminth infection is thought to drive the antibody response of malaria co-infected individuals towards the production of non-cytophilic subclasses ( IgG2 , IgG4 , and IgM ) , whereas protection against malaria is associated with the presence of the IgG1 and IgG3 cytophilic subclasses [11] . The cytokine milieu could also favour either pro- or anti- inflammatory reactions during malaria infection [12] . We recently conducted a randomized controlled trial of a T cell inducing vaccine in a malaria endemic area . 387 immunised children were monitored for episodes of malaria for 9 months after receiving either an experimental T cell inducing vaccine or a control ( rabies ) vaccination [13] . Since helminth infection might be a covariate for risk of febrile malaria , and might interfere with vaccine efficacy [11] , we prospectively obtained stool and urine samples for microscopy on all children enrolled into the trial , and measured peripheral eosinophilia . Here , we study the effect of helminth co-infections on the incidence of malaria .
405 children , aged 1 to 6 years ( inclusive ) , were randomized for either an experimental prime boost malaria vaccine or control vaccination ( rabies ) . The trial was assigned registration number ISRCTN88335123 with the International Standard Randomized Controlled Trial Number Register ( http://www . controlled-trials . com/isrctn/trial/ . The study was performed with the permission of KEMRI National Ethics Committees , and COREC , the NHS Central Office for Research Ethics Committees . The children were healthy , and resident in the study area . The study area was limited to Junju sub-location in Kilifi district , Kenya . The entomological inoculation rate ( EIR ) is 22–53 infective bites/person/year [14] . Malaria transmission continues all year , with two seasonal peaks . The process of randomization and vaccination is detailed elsewhere [13] . Of these 405 children , follow-up visits were completed for 387 children , among whom 122 children had one or more episodes of Plasmodium falciparum malaria . Treatment with mebendazole and praziquantel was available in local dispensaries , but drug use was monitored among the study participants for the duration of the study by locally-based field workers . Eosinophilia was measured in blood taken one week post-vaccination ( early May 2005 ) . Single urine and stool specimens collected for microscopy in January 2006 . Full blood counts were successfully analysed on 347 of the 387 children , and stool and urine samples collected for 315 and 294 children , respectively . Children were seen weekly by field workers , and blood films made when the temperature was ≥37 . 5°C . Field workers lived in the study area , and parents brought their children for assessment between regular weekly visits if the child developed fever . Treatment for episodes of malaria was with the Government of Kenya recommended first line treatment , artemether-lumefantrine . In analysis , a threshold of 2 , 500 asexual parasites of P falciparum per µl was used to determine febrile malaria . This threshold was derived in previous studies in Kilifi for use in children above one year of age [15] . Blood films were examined in duplicate by two microscopists , and examined a third time if there was a discrepancy . Eosinophil counts were measured by a COULTER® Ac·T™ 5diff CP . Wet microscopy was conducted on mid-morning terminal urine samples and stool . Children who were positive for schistosomiasis or gastrointestinal helminth infection were given praziquantel ( 40 mg per kg ) or mebendazole ( 100mg twice daily for 3 days ) , respectively . Vaccination was complete by day 90 , and treatment was given on day 150 . All analysis is conducted on pre-treatment data . The primary analysis was a log rank test comparing the time to the first or only episode of malaria between worm infected and uninfected children ( unadjusted ) . The hazard ratio and 95% confidence interval was also estimated by Cox's regression , adjusted for age group , ITN ( insecticide treated net ) use and village . Age group was a categorical variable with three levels ( 1–2 years old , 2–5 years old , 5–6 years old ) . Village had 5 levels . ITN use was defined as sleeping every night under a treated net , which had less than three holes into which a finger could comfortably fit . The vaccination was not efficacious [13] , and since adjusting for vaccine allocation made no difference to the results , the analysis presented here is unadjusted . Poisson regression was used to estimate the incidence rate ratio taking into account all malaria episodes , adjusted for the same covariates . A period of 28 days after each malaria episode was deducted from the person time at risk .
Stool samples were examined from 326 children . 25% were positive for ova indicating gastrointestinal worm infection . The three most common infections were Ascaris lumbricoides ( 18% ) , Trichuris trichiura ( 4 . 5% ) and hookworm ova ( 3% ) . The Kaplan Meier ( KM ) plot ( Figure 1 ) of clinical malaria episodes ( fever ≥37 . 5°C and parasitaemia >2 , 500 per µl ) shows similar rates for children with and without gastro-intestinal helminth infections ( p = 0 . 98 ) . Children with positive stool samples were treated soon after day 150 with mebendazole . Further modelling of the hazard and of multiple episodes ( Figure 2 ) excluded the period after treatment . In a Cox regression model , adjusted for age , village and ITN use , the modelled hazard ratio ( HR ) for first episode of malaria was 1 . 01 ( 95% CI 0 . 64–1 . 57 , p = 0 . 98 ) . In a Poisson regression model for the number of malaria episodes , the incidence rate ratio ( IRR ) was 0 . 83 ( 95% CI 0 . 51–1 . 33 , p = 0 . 45 ) . The risk associated with A . lumbricoides infection alone was also considered . The HR for first episode of malaria was 0 . 65 ( 95% CI 0 . 39–1 . 25 , p = 0 . 23 ) and in Poisson regression , the IRR was 0 . 67 ( 95% CI 0 . 37–1 . 2 , p = 0 . 18 ) . Urine samples were examined from 294 children . Only 24 children ( 8% ) had eggs of Schistosoma haematobium in urine , as identified by microscopy . The confidence intervals of estimates of clinical malaria risk among children with urinary schistosomiasis are therefore wide , and so data on the survival function is not shown in detail . The adjusted HR was 0 . 55 ( 95% CI 0 . 23–1 . 38 , p = 0 . 8 ) and the IRR was 0 . 69 ( 95% CI 0 . 28–1 . 73 , p = 0 . 2 . This is shown in Figure 2 . Schistosomiasis and gastrointestinal helminth infection were not associated ( χ = 0 . 53 , p = 0 . 46 ) and only 7 children had dual infections . Eosinophil counts were available from 347 children . 19% had highly elevated counts ( >1×106/ml ) and 31% had mildly elevated counts ( >0 . 5×106/ml ) . Peripheral blood eosinophil counts were conducted as a more sensitive marker of parasitic infection than microscopy . Neither Kaplan Meier plots ( Figure 1 , p = 0 . 71 ) nor adjusted estimates of risk ( Figure 2 ) showed significant variations in malaria episodes by eosinophil count . Mild eosinophilia was associated with a HR of 0 . 85 ( 95% CI 0 . 51–1 . 43 , p = 0 . 55 ) and an IRR of 0 . 86 ( 95% CI 0 . 51–1 . 4 , p = 0 . 6 ) . High eosinophilia was associated with a HR = 0 . 85 ( 95% CI 0 . 69–1 . 96 , p = 0 . 47 ) and an IRR of 0 . 98 ( 95% CI 0 . 63–1 . 5 , p = 0 . 57 ) . Schistosomiasis was seen in 0% of 1–2 year olds , 8% of 2–5 year olds and 14% of 5–7 year olds ( p = 0 . 02 ) , and gastrointestinal worm infections were seen in 11% of 1–2 year old , 25% of 2–5 year olds and 36% of 5–7 year olds ( p = 0 . 003 ) . Interestingly , eosinophilia became less frequent with age , from 63% of 1–2 year olds to 50% of 2–5 year olds and 42% of 5–7 year olds ( p = 0 . 03 ) . The average haemoglobin among children with gastrointestinal helminth infection was 10 . 47 g/dl ( 95% CI 10 . 2–10 . 8 ) and 10 . 21 g/dl ( 95% CI 10–10 . 4 ) among uninfected children , p = 0 . 23 . For children with schistovuria the average haemoglobin was 10 . 83 g/dl ( 95% CI 10 . 3–11 . 4 ) and 10 . 33 g/dl ( 95% CI 10 . 1–10 . 5 ) among uninfected children , p = 0 . 16 .
The data presented here were acquired during a trial designed primarily to examine vaccine efficacy . Children with gastrointestinal parasites or urinary schistosomiasis were treated with mebendazole or praziquantel , respectively . However , treatment was delayed for 150 days after screening , and the majority of malaria episodes ( 25% of children ) occurred before the children were treated . Only 6% of children had a malaria episode after treatment ( which coincided with the end of the high transmission season ) . We did not identify altered susceptibility to malaria associated with helminth infection , determined either by direct microscopy or inferred by peripheral eosinophilia . There were similar findings in Uganda [8] where soil-transmitted helminths were not associated with an increased risk of febrile malaria . In Senegal , other cohort studies have suggested that schistosomiasis increases the risk of febrile malaria [6] , and hospital-based case-control studies suggest both schistosomiasis and A . lumbricoides infection are risk factors for hospital admission with malaria [16] . However , a cohort study in Mali using weekly surveillance suggested that schistosome infection may actually protect from malaria [7] . Other studies suggesting that helminth infection might be protective relate to Thai adults , conducted by a group using case-control studies to examine severe disease [9] , [17]–[19] . However , this group also conducted one cohort study examining febrile disease , where susceptibility was conferred by intestinal helminth infection [20] . Recent reviews of the evidence appear to focus on the possibility that helminth infection might increase susceptibility to malaria [1] , [2] , [11] . Of the cohort studies , the sample sizes were 80 [5] , 511 [6] , 654 [7] and 435 [8] . The prevalence of helminth infection was 16% [5] , 67% [6] , 25% [7] and 17–47% [8] . Our study with 387 children , and prevalence of 25% , 50% and 8% for intestinal infection , eosinophilia and schistosomiasis , respectively , is therefore quite comparable in terms of power . Microscopy was only performed on a single specimen in our study , and we did not use concentration techniques . This would have missed lighter infections . This misclassification may have introduced bias against detecting an association . In previous studies that identified an increased risk of malaria in schistosome-infected children , multiple specimens were examined , although the greatest increase in risk was associated with heavier infections [6] , which are not likely to have been missed by a single examination . Studies that report a positive association with soil-transmitted helminths did not depend on examining multiple specimens [5] , [16] , and even in a low prevalence area with light infections , the yield from repeated examinations is low [21] . The use of artesunate further complicates our assessment . This appears to result in a reduction in a relatively rapid reduction in egg production among infected children [22] , and it is unclear how rapidly this might alter host susceptibility to malaria . Our primary analysis of gastrointestinal infection did not differentiate on the basis of species , but secondary analysis that was restricted to A . lumbricoides infection did not show a tendency towards increased risk of malaria in infected children either . The low prevalence of helminth infection limits the power of the study , but by how much ? This can be judged by the confidence intervals associated with the estimates of efficacy , from power calculations , and from inspecting the KM plots . The CIs around the estimates of risk are narrower than a halving or doubling of risk in each analysis conducted . The narrowest CIs are those derived for all gastrointestinal helminth infections and eosinophilia , followed by A . lumbricoides infection alone , with CIs for Schistosoma haematobium the widest . The tendency is in the direction of decreased rather than increased susceptibility , so the possible increases in risk that could have been missed are not large for gastrointestinal helminth infections and eosinophilia , but quite large effects of schistosome infection might have been missed . The primary analysis was log rank testing of time to event . Assuming a constant effect over time , with the prevalence of gastrointestinal helminth infection at 25% , then a log rank analysis had 80% power to detect a 70% increase or 50% decrease in risk due to helminth infection . Power was very low for schistosomiasis , with 80% power to detect a 110% increase or 90% decrease , but 80% power to detect an 80% increase or 60% decrease for A . lumbricoides infection alone . However , as it turned out , the upper limits of the confidence intervals measured suggest that 60–80% increases in risk are unlikely , although greater decreases in risk have not been ruled out . Previous studies have suggested a doubling in risk of febrile malaria in schistosome-infected children [6] , and 64% from soil-transmitted helminths [5] . The increased risk seen secondary to schistosomiasis was seen only in the heavily infected children , and our study did not differentiate heavy from light infections . On inspection of the KM plots , there is no marked separation of the helminth-infected and uninfected at any stage . Between 50 and 125 days , helminth-infected children have acquired slightly fewer cases of febrile malaria , but this difference is slight , and the lines converge by 150 days . After 150 days there is a slight reduction in the number of children with febrile malaria among those with moderate and high eosinophil counts . This suggests that further studies to examine the impact of helminth infection in this population would need to be considerably larger . The results of individual studies may be dependant on the transmission intensity of both worm and Plasmoidium infections . In our study , the moderately high malaria transmission meant that significant immunity had been acquired by children aged 5–7 compared with 1–2 year olds ( HR 0 . 24 , 95% CI 0 . 1–0 . 54 , p = 0 . 001 ) . However , worm infection is more frequent in the older age groups , and the interaction between worms and malaria may depend critically on being exposed to helminth infections while immunity to malaria is being acquired . We did not measure intensity of infection in this study , and it is possible that among older children , with more intense worm infections , there might have been a relationship between helminth infection and malaria . However , these children would have a low incidence of malaria . Previous studies identifying increased susceptibility to malaria among helminth infection have been conducted in settings of more prevalent helminth infection . 50% of children had either mild or high eosinophilia , compared with a prevalence of 7% for urinary schistosomiasis and 25% for gastrointestinal helminth infection . The higher prevalence may reflect infections missed by microscopy , but also could be due to filarial infection ( which we did not examine for ) . Strongyloides stercoralis infection is also associated with eosinophilia , and would not have been identified by the stool examination conducted in this study . Eosinophilia is closely associated with the TH2 cytokine responses which might result in the increased susceptibility to malaria [23] , and might therefore have been expected to be a more discriminatory marker . Eosinophilia was not associated with altered rates of malaria in this study . However , the potentially deleterious responses that can be induced by helminths include TH1 and regulatory responses , of which eosinophilia would not be a marker . We found that eosinophilia became less prevalent with age , despite the prevalence of helminth infection rising , suggesting that eosinophilia becomes suppressed with chronic helminth infection . As well as immunological mechanisms [24] , [25] , it has been hypothesised that helminth induced anaemia induces susceptibility to malaria [24] , although other data suggest that iron deficiency protects against malaria [26] . In any case , helminth infected children in our study were not significantly more anaemic . This study suggests that in a cohort of one to six year old children in a moderately high malaria transmission area are not at increased risk of malaria from concurrent helminth infection , as evidenced either by microscopy or by eosinophilia . In our study , the prevalence on helminth infection was relatively low , and heavy infections were probably infrequent . This reflects the age of the children enrolled . However , it would not have been practical to study older children at this transmission intensity , since immunity against febrile disease is acquired rapidly ( the HR among five to six year old children was 0 . 29 ( 95% CI 0 . 15–0 . 58 ) . In areas with lower malaria transmission , older children may retain susceptibility to malaria . What further studies should be conducted ? An interaction between helminth infection and malaria is of considerable public health importance . If present only at lower transmission intensities , this would still account for a very significant number of infections globally [27] , and even a slight impact on high transmission areas might still be significant [2] . That we have not identified an effect in the present study does not exclude a significant effect , given the caveats associated with the methodology and sample size . Further studies will need more detailed assessments of individual children as well as a greater sample size . To overcome the caveats identified in this study , multiple examinations of quantified stool specimens will be required during the period of follow-up , multiple geographical areas and a wider age group should be included , and attention should be given to filarial infection , Strongyloides stercoralis and eosinophilia . Only a large and detailed study of this nature will satisfy concerns about multiple infections , changing infection status of individual children , and increased susceptibility depending critically on worm burden and age of acquisition of immunity to malaria . | Malaria infection and other parasitic infections are widespread in developing countries . There is evidence from some studies that intestinal worm infections may increase the risk of developing febrile malaria . However , the evidence is mixed , and some studies have found no effect or even protective effects . A vaccine trial was recently conducted to assess the efficacy of a candidate malaria vaccine . Episodes of malaria were monitored . The vaccine was not protective , but data was also recorded on the prevalence of worm infections . The rates of febrile malaria did not seem to vary according to worm infection in this study . However , because of the relatively low prevalence of worm infection , the study did not have high power . Given the conflicting findings in the literature , and the potential for the effect of worm infection to vary geographically , it is important that larger , definitive studies are conducted , since even quite small effects might be important for global public health . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases"
] | 2008 | Helminth Infection and Eosinophilia and the Risk of Plasmodium falciparum Malaria in 1- to 6-Year-Old Children in a Malaria Endemic Area |
Infectious Leptospira colonize the kidneys of reservoir ( e . g . rats ) and accidental hosts such as humans . The renal response to persistent leptospiral colonization , as measured by urinary protein biosignatures , has not been systematically studied . Urinary exosomes--bioactive membrane-bound nanovesicles--contain cell-state specific cargo that additively reflect formation all along the nephron . We hypothesized that Leptospira-infection will alter the content of urine exosomes , and further , that these Leptospira-induced alterations will hold clues to unravel novel pathways related to bacterial-host interactions . Exosome protein content from 24 hour urine samples of Leptospira-infected rats was compared with that of uninfected rats using SDS-PAGE and liquid chromatography/tandem mass spectrometry ( LC-MS/MS ) . Statistical models were used to identify significantly dysregulated proteins in Leptospira-infected and uninfected rat urine exosomes . In all , 842 proteins were identified by LC-MS/MS proteomics of total rat urine and 204 proteins associated specifically with exosomes . Multivariate analysis showed that 25 proteins significantly discriminated between uninfected control and infected rats . Alanyl ( membrane ) aminopeptidase , also known as CD13 topped this list with the highest score , a finding we validated by Western immunoblotting . Whole urine analysis showed Tamm-Horsfall protein level reduction in the infected rat urine . Total urine and exosome proteins were significantly different in male vs . female infected rats . We identified exosome-associated renal tubule-specific responses to Leptospira infection in a rat chronic colonization model . Quantitative differences in infected male and female rat urine exosome proteins vs . uninfected controls suggest that urine exosome analysis identifies important differences in kidney function that may be of clinical and pathological significance .
Leptospirosis is among the world’s most important zoonotic infectious diseases [1 , 2] , characterized by variable manifestations ranging from asymptomatic or self-resolving acute febrile illness to severe disease with a combination of fever , acute kidney injury , jaundice , severe pulmonary hemorrhage syndrome , refractory shock , and aseptic meningitis [3] . Important advances have been made in diverse aspects of leptospirosis including the differential host responses to leptospiral infection [4–9] . Despite these advances , mechanistic details by which end organ damage develops in some individuals but not in others remain to be elucidated . Further , factors that govern host susceptibility to leptospiral infection are not well understood . A recent study by our collaborative group found that approximately 6% of randomly sampled individuals in a highly endemic rural Amazonian village were chronically colonized by Leptospira without any recent clinical evidence of infection [10] . Although renal colonization by leptospires may occur in humans without serological or clinical evidence of infection , the clinical relevance and functional consequences of leptospiral colonization in humans remain to be characterized . New , less invasive , less expensive and more practical tools ( compared to kidney biopsy ) are needed to study pathologic changes in the kidney in order to understand clinically relevant sequelae of infection . In this report , we use proteomic analysis of urine exosomes in a rat chronic colonization model as a non-invasive window to kidney function in asymptomatic leptospiral infection . Exosomes are nanovesicles that are released from cells as a mechanism of intercellular communication [11] . Characterization of exosomes from different biological samples has shown the presence of common as well as cell-type specific proteins . The protein content of exosomes has been shown to be modified under pathological or stress conditions [12–14] . Since exosome contents are specifically derived from cellular components , here we tested the hypothesis that urinary exosome protein content from a rat infected with Leptospira would be different from that of the uninfected rat , and that these differences would hold key information about the pathways mediating host responses to Leptospira infection . After renal colonization , persistent shedding of Leptospira is clearly established in carrier animal hosts , especially rodents . However , they rarely develop symptoms and are not noticeably impaired by infection of their kidneys [1 , 15 , 16] . We have recently detected chronic asymptomatic renal colonization by Leptospira in human subjects from a rural Amazonian village [10] . We reasoned that the structural and functional changes in the kidney that arise following asymptomatic Leptospira infection are different from symptomatic disease . These differences between the asymptomatic and symptomatic leptospirotic kidneys can be understood by studying the downstream products of the kidney , such as urine . Given the nephron cell-state-specific cargo of the urinary exosome , we hypothesized that urine exosome analysis holds key information that is relevant to differences between clinically symptomatic and asymptomatic leptospirosis infection . This report is the first step towards testing this hypothesis . Here we report our preliminary findings from the exosome proteomic analyses of urines from rats infected with Leptospira using uninfected rat urine exosome as controls . We also studied the host-response to Leptospira infection in male and female rats separately .
This work was approved by the Institutional Animal Care and Use Committee of the University of California , San Diego . Three-week old Sprague Dawley rats ( Charles River Laboratories , USA ) ( 6 male rats and 6 female rats ) were housed in cages of 1–2 rats each , with food and water provided ad libidum . Six animals ( 3 males and 3 females ) were inoculated intraperitoneally with 108 mid-log phase L . interrogans serovar Copenhageni strain HAI1026 . Uninfected controls were inoculated ip with sterile EMJH . To determine the health status , weight , general body condition , posture , activity , appetite food and water consumption of each animal was monitored daily . Infection was confirmed by serology ( microscopic agglutination test ) and quantitative polymerase chain reaction ( qPCR ) of weekly urines , and Warthin-Starry silver staining of kidney sections following necropsy . In order to keep the minimum number of animals for the purpose of statistical calculations , we used this small number of animals . This is also discussed as a limitation of the study under discussion . Urine from each animal was collected weekly starting 7 days after post-challenge by placing the animals individually for 20–24 hrs in metabolic cages , and was captured into containers containing Roche Complete Protease Inhibitor , one tablet per 5 mL urine . Urines were separately centrifuged at 3000 x g for 30 min . The supernatant was withdrawn , the pH adjusted to 7 , aliquoted and frozen at -70°C until further analysis . Genomic DNA was extracted from these weekly urines and qPCR was used to assess leptospiruria . In addition , at necropsy , kidneys were harvested , fixed in formalin , paraffin embedded , and infection confirmed by Warthin-Starry silver stain ( S1 and S2 Figs . ) . Exosomes from terminal urine samples from infected and uninfected rats were prepared using an in-house protocol developed based on the solvent exclusion principle using polyethylene glycol ( PEG ) -induced precipitation . To prevent naturally occurring peptides in the exosome from confounding post in-gel trypsinization peptide information of the full-length proteins , we conducted 1 dimensional SDS-PAGE of the exosome proteins prior to in-gel trypsinization . Each rat urine sample ( only terminal sample ) was separately analyzed , without pooling any sample . Each rat urine sample was run in a separate gel lane . Gel slices for each lane were cut to 1 mm x 1 mm cubes and destained 3 times by first washing with 100 μL of 100 mM ammonium bicarbonate for 15 min , followed by addition of the same volume of acetonitrile ( ACN ) for 15 min [17] . The supernatant was transferred to a clean tube and samples lyophilized and reduced by mixing with 200 μL of 100 mM ammonium bicarbonate-10 mM DTT then incubated at 56°C for 30 min . The liquid was removed and 200 μL of 100 mM ammonium bicarbonate-55mM iodoacetamide was added to gel pieces , which were then incubated at room temperature in the dark for 20 min . After removal of the supernatant and one wash with 100 mM ammonium bicarbonate for 15 min , an equal volume of ACN was added to dehydrate the gel pieces . The solution was then removed and samples lyophilized . For digestion , ice-cold trypsin ( 0 . 01 μg/μL ) in 50 mM ammonium bicarbonate solution was added in enough amounts to cover the gel pieces and set on ice for 30 min . After complete rehydration , the excess trypsin solution was removed , replaced with fresh 50 mM ammonium bicarbonate , and left overnight at 37°C . The peptides were extracted twice by the addition of 50 μl of 0 . 2% formic acid and 5% ACN and vortexed at room temperature for 30 min . The supernatant was removed and saved . A total of 50 μL of 50% ACN-0 . 2% formic acid was added to the sample , which was vortexed again at room temperature for 30 min . The supernatant was removed and combined with the supernatant from the first extraction . The combined extractions from all the gel slices in a lane pertaining to a single rat urine exosome sample was separately analyzed directly by liquid chromatography ( LC ) in combination with tandem mass spectroscopy ( MS/MS ) using electrospray ionization . Thus , multiple slices of gels representing a single rat urine exosome sample was used for mass spectrometry . Two replicates per rat exosome sample were run on the MS . Trypsin-digested mixtures were analyzed by the Eksigent nanoLC-Ultra 2D System ( Eksigent , AB SCIEX Dublin , CA , USA ) combined with cHiPLC-nanoflex system ( Eksigent ) in trap-elute mode . Briefly , samples were first loaded on the cHiPLC trap ( 200 μm x 500 μm ChromXP C18-CL , 3 μm , 120 Å ) and washed in isocratic mode with 0 . 1% aqueous formic acid for 10 min at a flow rate of 3 μL/min . The automatic switching of cHiPLC ten-port valve then eluted the trapped mixture on a nano cHiPLC column ( 75 μm x 15 cm ChromXP C18-CL , 3 μm , 120 Å ) , through a 45 min gradient of 5–50% of eluent B ( eluent A , 0 . 1% formic acid in water; eluent B , 0 . 1% formic acid in acetonitrile ) , at a flow rate of 300 nL/min . To preserve system stability , in terms of elution times of components , trap and column were maintained at 35°C . Mass spectra were acquired using a QExactive mass spectrometer ( Thermo Fisher Scientific , San José , CA , USA ) , equipped with a nanospray ionization source ( Thermo Fisher ) . Nanospray was achieved using a coated fused silica emitter ( New Objective , Woburn , MA , USA ) ( 360 μm o . d . /50 μm i . d . ; 730 μm tip i . d . ) held at 1 . 5 kV . The ion transfer capillary was held at 220°C . Full mass spectra were recorded in positive ion mode over a 400–1600 m/z range and with a resolution setting of 70000 FWHM ( @ m/z 200 ) with 1 microscan per sec . Each full scan was followed by 7 MS/MS events , acquired at a resolution of 17 , 500 FWHM , sequentially generated in a data dependent manner on the top seven most abundant isotope patterns with charge ≥2 , selected with an isolation window of 2 m/z for the survey scan , fragmented by higher energy collisional dissociation ( HCD ) with normalized collision energies of 30 and dynamically excluded for 30 s . The maximum ion injection times for the survey scan and the MS/MS scans were 50 and 200 ms and the ion target values were set at 106 and 105 , respectively . All data generated were searched using the Sequest search engine contained in the Thermo Scientific Proteome Discoverer software , version 1 . 4 . The experimental MS/MS spectra were correlated to tryptic peptide sequences by comparison with the theoretical mass spectra obtained by in silico digestion of the Rattus norvegicus protein database downloaded January 2013 from the National Centre for Biotechnology Information ( NCBI ) website ( www . ncbi . nlm . nih . gov ) . The following criteria were used for the identification of peptide sequences and related proteins: trypsin as enzyme; three missed cleavages per peptide were allowed and mass tolerances of ± 50 ppm for precursor ions and ± 0 . 8 Da for fragment ions were used . Validation based on separate target and decoy searches and subsequent calculation of classical score-based false discovery rates ( FDR ) were used for assessing the statistical significance of the identifications . Finally , to assign a final score to proteins , the SEQUEST output data were filtered as follows: 1 , 5; 2 . 0; 2 . 25 and 2 . 5 were chosen as minimum values of correlation score ( Xcorr ) for single-; double-; triple- and quadrupole-charged ions , respectively . A high stringency was guaranteed using parameters previously described [18] and the false-positive peptide ratio , calculated through a reverse database , was less than 3% . The output data , protein lists , obtained from the Sequest search were compared using an in-house software , namely , the Multidimensional Algorithm Protein Map ( MAProMa ) [19] . The relative abundance of polypeptides was calculated from the normalized spectral abundance factor ( NSAF ) using the method of Paoletti et al [20] taking into consideration the number of peptides as well as the length of the polypeptide contributing to their respective abundance . To enable comparison of samples from a subject across different time points or across different groups , each animal’s total proteome was normalized to 1 . Subsequently the relative contribution of each protein from a given animal from a group was expressed as percentage of the total . Peptide numbers corresponding to a protein were thus more of raw data nature whereas the NSAF number included the peptide number as well as the total length of the protein . The peptide counts data were log-transformed prior to analysis by multivariate partial least squares discriminant analysis ( PLSDA ) , and univariate 1-way ANOVA with unpaired comparisons , Variable Importance in Projection analysis and post hoc correction by Wilcoxon Rank test in MetaboAnalyst [21] . In all analyses , p ≤ 0 . 05 was considered statistically significant . Analyses including Student’s t-tests , Partial-Least Squares Discriminant Analysis ( PLS-DA ) and variable importance in projection ( VIP ) were performed with the MetaboAnalyst 2 . 0 web portal ( www . metaboanalyst . ca ) [21] . To reduce systematic variance and to improve the performance for downstream statistical analysis normalization and transformation of raw data were performed before the t-tests , PLS-DA and VIP analysis . Normalization by sum of the spectral count as mentioned previously was used to overcome the variance between the analyzed samples . To make each feature comparable in magnitude to each other , data were transformed by taking the natural log of the concentration values of the analyzed proteins . The data were additionally auto-scaled ( mean-centered and divided by the standard deviation of each variable ) . Univariate analysis was used to check the differences in the concentrations of the analyzed exosome protein spectral count between the control and infected rat urine samples . The infected rat urines were also split into male rat and female rat categories . Paired Student’s t-test was applied to examine each variable ( ratio of individual protein to total concentration in each group considered ) . PLS-DA and VIP were used both for the classification and significant feature selection . A VIP plot , which is commonly used in PLS-DA , ranks proteins based on their importance in discrimination between the urinary exosomes from infected and the uninfected rats . The VIP score is a weighted sum of squares of the PLS loadings . The amount of explained Y-variance in each dimension influenced the weights [22] . Protein candidates with a false discovery rate ( FDR ) of ≤10% were qualified for subsequent validation by Western immunoblotting . Antibody against alanyl aminopeptidase was purchased from Proteintech Group , Inc . , ( Chicago , IL , USA ) . The THP antibody was from Sigma Chemical Co ( St . Louis , MO , USA ) . HRP-conjugated secondary antibody was from GE Life Sciences ( Piscataway , NJ , USA ) . SDS-PAGE gels ( with 10% acrylamide ) were used to resolve 100 μg of protein either from exosomes or total urines of male and female rats infected with L . interrogans serovar Copenhageni . After separation the proteins were transferred to nitrocellulose paper , blocked , and incubated with primary antibody overnight before washing with Tris-buffered saline , incubation for 1 h with HRP-secondary antibody conjugate and visualized by developing as described in previous publications from our laboratory [23 , 24] . The quantification of the Western immunoblot bands was performed using Image J software ( NIH ) as previously described [24] , and plotted using Graphpad Prism software ( San Diego , CA , USA ) .
Rat urine samples were analyzed for overall protein identification by a combination of SDS-PAGE and mass-spectrometry . We found that the infected rat urine shows an overwhelming increase in both quality and quantity of the protein content , as shown in Panel A of Fig . 1 ( 156 versus 503 proteins unique to uninfected versus infected urines ) . A total of 842 proteins were detected , with further classification of subgroups in the infected animals based on gender as shown in Panel B of Fig . 1 . In total , 842 proteins were identified from the total urine animals with distribution as shown in Fig . 1 Panel A and B ) . Importantly , 180 proteins and 272 proteins were unique to the urines of female and male rats infected with L . interrogans serovar Copenhageni respectively , as compared to 156 proteins unique to the uninfected rat urines . These differences in the composition and quantity of proteins from infected and uninfected rats potentially indicates the reactions induced in these animals by leptospiral infection . The identity of each of these proteins is as given in S1 Table . Given that the urine exosomes reflect intracellular milieu of all types of various cells lining the nephron in kidney , and emerging evidence from literature that kidney functional alterations are induced due to leptospiral infection , we next focused on exosomes . We found that a total of 204 exosome proteins were identified classifiable into 7 different groups as noted in S2a-S2g Table and summarized in the Venn diagram ( Panel C , Fig . 1 ) . The exosome protein constitution also showed increase in the number of proteins expressed in the exosome , viz 32 proteins uniquely present in the uninfected exosome versus 57 unique proteins in the infected rat exosome . We further conducted the multivariate partial least squares-discriminant analysis ( PLS-DA ) on these proteins . The analysis depicted in Fig . 2 shows clear separation between uninfected and infected rat urine exosome protein content , suggesting the differences between uninfected and infected rats . Urinary exosome proteins in male vs . female infected rat urine were different as determined by PLS-discriminant analysis ( Fig . 3 ) . Moreover , the infected male rat urine exosome contents were far more different from those of uninfected rats compared to the infected female rat urine exosome contents . The VIP ( Variable Importance in Projection ) score of 25 proteins was higher than 1 . 5 ( Fig . 4 and Table 1 ) . Qualitatively , a total of 57 proteins were present among all the infected rats . Of these , only 3 were shared between infected males and females , while 37 were unique to infected males and 17 were unique to infected females . Further , we conducted separate analyses of proteins between proteins of proteins of male infected and female infected rats . Table 2 depicts male infected versus control rat urine exosome analysis . Accordingly , 11 proteins were significantly altered ( p < 0 . 05 ) . Table 3 depicts female infected versus control rat urine exoosme analysis , according to which the number of significantly dysregulated proteins was 9 . In the male infected rat urine exosome , the alanyl ( membrane ) aminopeptidase upregulation not only reached the highest level of significance ( p = 0 . 00019 ) but also had the lower FDR ( 3 . 22% ) . In the female infected rat however , although this upregulation was significant compared to the uninfected rat , the FDR did not reach the cutoff of <10% ( 14 . 37% ) . Thus both qualitatively and quantitatively , the protein content of exosomes showed gender specificity in infected rats . Of interest , the top discriminator between control and infected rats , namely alanyl aminopeptidase of the membrane origin , is also known as aminopeptidase neutral ( APN ) or CD13 [25 , 26] . Furthermore , CD13 also shows different levels of dysregulation between infected male and infected female rats ( Table 1 ) . A separate analysis of the proteins dysregulated between control and infected rats showed that 11 proteins were significantly different in urinary exosomes ( p <0 . 05 , Table 4 ) . However , only one protein had an FDR of <10% , namely alanyl ( membrane ) aminopeptidase , also known as CD13 . By both multivariate analysis of PLS-DA ( Fig . 4 ) and univariate ANOVA , CD13 is significantly upregulated in the infected rat urine exosome . When the analysis was conducted without gender specificity ( Table 4 ) , only CD13 showed an FDR of <10% among these 11 dysregulated proteins . Taking into account gender specificity ( Table 2 ) CD13 was not only significantly upregulated , but also had an FDR of <10% . However , in females , this protein was only significantly upregulated but did not reach the FDR cutoff ( Table 3 ) . The VIP scores of proteins identified in the rat urine exosomes shows CD13 to be a top discriminant between infected and uninfected rat urine exosome ( Table 1 ) . Twenty-five proteins had a VIP score of >1 . 5 , fulfilling criteria to classify a protein as a reliable discriminant . Our analysis shows that this value 5 . 72 for CD13 . Further , Western immunoblotting of the protein showed that the exosome content of CD13 closely tracked the proteomic data , with a slight increase in the infected female urine exosome , and robust increase in the infected male urine exosome ( Fig . 5a ) . The observed difference was significant ( Fig . 5b ) . Given the tubular location of CD13 , we tested the hypothesis that other proteins reflecting the tubular function may be affected in response to leptospiral infection . We chose to study the most abundant protein in the normal urine , namely the Tamm-Horsfall Protein ( THP ) . Western immunoblotting of THP in urine of infected and uninfected rats showed robust THP expression in the uninfected rats and significantly lower THP in infected male and female rats , supporting this hypothesis ( Fig . 6 ) .
Many conclusions can be drawn from this study: Although the general quantity of proteins in total urine and exosome component show a similar trend , qualitatively the proteins are different based on identifications , the GI numbers and hence the pathways they belong to . Utility of exosome Analysis in characterizing a complex biochemical phenomenon such as the host-response to leptospirosis infection is very high , and has many implications: Public health interest: given that nephron-cell specific cargo that urine exosome carries , primarily the pathway employed or dysregulated for infecting the host to bring about the infection phenotype , and secondly , the response that is unique to each host studied potentially opens up the possibilities of disease-specific and severity-specific treatment to leptospirosis . Non-invasive: it would be ideal to biopsy every infected individual to understand the various facets of leptospirosis infection in order that our knowledge about Leptospira increases . However , it is neither practical nor ethical , given the invasive nature . Urine exosome analyses is non-invasive , highly specific and provides a window into the organ level structural or functional changes . If implemented quickly enough in a translational research setting , this specificity potentially imparts ability to move the field towards personalized medicine . Gender-specific host responses to disease onset: our data shows that this methodology of studying differences between male and female host response to infection can be applied in the setting of human disease . Although we should expect different molecules other than CD13 to by dysregulated , the individual-specific pathways mediating different magnitude responses to the same infection can be accurately measured using exosome markers that are either surrogately or directly linked to a particular host-response event . This assumes importance especially in the setting of sub-clinical leptospirosis infection where in the patient does not display any clinical features/symptoms but continues to play host to the bacterium . | Leptospirosis is a bacterial disease commonly transmitted from animals to humans . Though this disease affects more than three quarters of a million people every year and takes a disproportionate toll on the poor in in tropical regions , few virulence factors have been identified and very little is known regarding the pathogenesis of leptospirosis . Symptoms vary from fever and fatigue to severe pulmonary hemorrhage and death . Approximately 5–10% of Leptospira infections in humans are chronic ( >1 year ) and asymptomatic ( no overt signs of disease ) . Nonetheless , very little is known about the clinical significance of these infections . In this report , we show that non-invasive tools namely proteomic analysis of urinary exosomes can be used to identify differences between healthy and Leptospira-infected rat kidney and between Leptospira-infected male and female rat kidney . In future studies , these analyses will be extended to determine clinical significance and extent of renal dysfunction in the asymptomatic human . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Proteomic Analysis of Urine Exosomes Reveals Renal Tubule Response to Leptospiral Colonization in Experimentally Infected Rats |
At the interface between T cell and antigen-presenting cell ( APC ) , peptide antigen presented by MHC ( pMHC ) binds to the T cell receptor ( TCR ) and initiates signaling . The mechanism of TCR signal initiation , or triggering , remains unclear . An interesting aspect of this puzzle is that although soluble agonist pMHCs cannot trigger TCR even at high concentrations , the same ligands trigger TCR very efficiently on the surface of APCs . Here , using lipid bilayers or plastic-based artificial APCs with defined components , we identify the critical APC-associated factors that confer agonist pMHCs with such potency . We found that CD4+ T cells are triggered by very low numbers of monomeric agonist pMHCs anchored on fluid lipid bilayers or fixed plastic surfaces , in the absence of any other APC surface molecules . Importantly , on bilayers , plastic surfaces , or real APCs , endogenous pMHCs did not enhance TCR triggering . TCR triggering , however , critically depended upon the adhesiveness of the surface and an intact T cell actin cytoskeleton . Based on these observations , we propose the receptor deformation model of TCR triggering to explain the remarkable sensitivity and specificity of TCR triggering .
Using T cell receptors ( TCRs ) as sensors , T cells probe the surface of antigen-presenting cells ( APCs ) for the presence of antigenic ( agonist ) peptides presented by major histocompatibility complex ( pMHC ) molecules . Engagement of TCRs by agonist pMHCs initiates a signal , which is transmitted to the nucleus via kinase cascades and protein translocation , leading , ultimately , to T cell activation . In clear contrast to advances in our understanding of intracellular TCR signaling pathways , the question of how binding of pMHC initiates or triggers TCR signaling remains unclear [1] . This is despite knowledge gained from the study of triggering mechanisms of other receptor systems , for example , the conformational change of G protein-coupled receptors and the dimerization of growth factor receptors . The difficulty in resolving the mechanism of TCR triggering is attributed to the complexity of multichain TCR/CD3 structure , the diversity of peptides presented by MHCs on APCs , and importantly , the complex environment where pMHC-TCR interaction takes place . One fundamental question still unanswered is the minimum requirements for TCR triggering . In terms of the pMHC ligand , a critical question is whether a monomeric agonist pMHC alone can trigger TCR independently , or whether it must act cooperatively with another neighboring pMHC , either agonist or endogenous , as a dimer to simultaneously engage , and in effect crosslink , two TCRs . In terms of the environment or the context of TCR triggering , critical questions are whether soluble agonist pMHCs are capable of triggering TCR or if they must be surface-anchored , whether molecules other than pMHCs on APCs , e . g . , costimulatory molecules , contribute to TCR triggering , and whether and how physical and mechanical aspects of the T cell–APC interaction play a role in TCR triggering . There are reports showing that TCR is no different from receptors for soluble ligands , e . g . , hormone receptors , and that soluble monomeric pMHCs are sufficient to trigger TCR [2–4] . These data suggested that TCR must be triggered through either TCR conformational change or heterodimerization of TCR and CD4 or CD8 coreceptor via simultaneous pMHC binding . In these studies , however , binding of peptides [5] or pMHC molecules to the surface of the cell or culture dish was difficult to rule out . Mechanistically , with one exception [6] , rather extensive crystallography studies have not revealed global large-scale conformational changes of TCR after engagement of agonistic pMHCs [7–14] . It is also uncertain whether coreceptors play a role in TCR triggering , since TCR triggering in vitro and in vivo [15 , 16] can occur in the complete absence of CD4 or CD8 . Moreover , these studies are contradicted by data showing the total lack of TCR triggering capability of soluble agonist pMHCs in solution even at very high concentrations [17–22] . In sharp contrast , however , many studies have demonstrated the remarkable potency of a few agonist pMHCs on APCs to trigger TCR [23–25] . If TCR triggering by agonist pMHCs depends upon the context of the APC , then the question becomes what else at the interface of T cell–APC interaction contributes to TCR triggering . Based upon observations of T cell activation by antibody crosslinking , TCR crosslinking by agonist–agonist pMHCs has been proposed as the mechanism of TCR triggering . This theory relies upon the presence of putative agonist–agonist pMHCs on the APC surface , and is supported by observations that a significant fraction of MHCs on APCs are immobile [26 , 27]; thus , two agonist pMHC monomers could act as a dimer if they are immobilized close to each other . However , since TCR can be triggered by only a few agonist pMHCs amongst a huge number of endogenous pMHCs on the APC , the chance of agonist–agonist pMHC formation should be very small . Getting around this issue is a recent theory that TCR is triggered by a “pseudodimer” of an agonist pMHC and an endogenous pMHC [28 , 29] , based upon data showing that some endogenous pMHCs enhance TCR triggering by agonist pMHCs . In addition to the possible contribution of endogenous pMHCs on APCs , van der Merwe and colleagues have postulated that the two-dimensional ( 2D ) nature of the pMHC–TCR interaction plays a critical role in triggering . The proposed kinetic-segregation model [30–32] suggests that TCR signal initiation is induced by segregation of small kinase-associating molecules and large phosphatase-associating molecules at the tight junction between a T cell and an APC , leading to a net increase in tyrosine kinase activity proximate to the TCR/CD3 complex . This model would obviate the need for receptor crosslinking and require only monomeric agonist pMHC . With conflicting data and the multiple models proposed , it is clear that there is no consensus regarding the mechanism of TCR triggering . A solution to this puzzle would be aided by definition of the critical factors involved in the T cell–APC interaction that directly contribute to TCR triggering by agonist pMHCs . Here , we test TCR triggering using artificial APCs consisting of fluid lipid bilayers or fixed plastic surfaces with defined components . We demonstrate that TCR can be triggered by a very low number ( 1–10 ) of monomeric agonist pMHCs . TCR triggering is independent of endogenous pMHCs and of any other molecule found on real APC surfaces , but is critically dependent upon ( 1 ) surface-anchoring of pMHC , ( 2 ) T cell adhesion to the surface , and ( 3 ) intact actin cytoskeletal function . Based upon these data , and incorporating the impact of mechanical stress on pMHC-TCR binding kinetics , we propose the receptor deformation model of TCR triggering by monomeric pMHCs on a surface . In this model , TCR is triggered by TCR/CD3 conformational changes induced by a cytoskeletal pulling force , transferred via specific pMHC-TCR interactions with sufficient resistance to rupture under force .
To anchor pMHCs on a fluid surface , we developed a planar lipid bilayer-based artificial APC ( Figure 1A ) . This consisted of a glass-supported lipid bilayer anchored with pMHC and ICAM-1 , via interactions between nitrilo triacetic acid-Ni ( NTA-Ni ) lipid head groups on the bilayer and 6× histidine ( HisTag ) at the membrane-proximal ends of the proteins . An initial consideration was that artificial pMHC dimers may be formed by monomers immobilized in close proximity on lipid bilayers with defects in fluidity or homogeneity . Avoidance of such de facto dimer formation was crucial for interpretation of our experimental data . To this end , a novel technique of lipid bilayer expansion was employed to achieve a higher level of large-scale lipid bilayer homogeneity and fluidity compared to previously reported methods . The newly expanded area of lipid bilayer showed excellent homogeneity and stability in buffer containing bovine serum albumin ( BSA ) , with no defects visible by fluorescence microscopy ( Figure 1B ) . As expected , given its genesis by lipid translocation , the newly expanded lipid bilayer showed a high level of fluorescence recovery after photobleaching ( FRAP ) of a spot of 4-μm radius , indicating a mobile fraction of close to 100% and a calculated diffusion coefficient of approximately 0 . 8 μm2/sec ( Figure 1C ) , in agreement with typical lipid diffusion rates in biological membranes [33] . The lipids diffuse freely over long range as shown in the recovery of a larger bleaching spot ( Figure S1A ) . Only this newly expanded area of lipid bilayer was used for interaction with T cells . To anchor monomeric pMHC on the bilayer , a lipid with an NTA head group ( DOGS-NTA ) was doped in the bilayer at 5 mol% and charged with Ni2+ to allow binding of soluble ligands with HisTags . To generate soluble monomeric pMHCs , extracellular domains of MHC class II IEk molecules , each with one of three different covalently linked peptides , were expressed as secreted forms using a baculovirus system in insect cells [34]: IEk-MCC ( agonist peptide moth cytochrome c residues 88–103 ) [35] , IEk-HSP70 ( endogenous peptide HSP70 234–248 ) [29 , 36] , and IEk-ER60 ( endogenous peptide ER60 448–461 ) [29 , 37] . IEk-MCC is a well-characterized ligand for the Vα11Vβ3 TCR expressed on CD4+ T cells from the TCR transgenic mice AD10 , AND , and 5C . C7 [38–40] . All IEk proteins were engineered to have an AviTag sequence followed by a HisTag sequence connected to the membrane-proximal ends of the β chain via flexible linkers . The AviTag sequence allows biotinylation of a single lysine residue by the BirA enzyme . All proteins displayed the same single peak corresponding to approximately 50 kDa in gel filtration chromatography ( Figure S2A ) . Their secondary structures were confirmed by circular dichroism spectra ( Figure S2B ) and thermal melting profiles ( Figure S2C ) , both of which are consistent with that of IEk with bound MCC peptide , described previously [41] . In particular , all three proteins showed similar thermal melting profiles with sharp transitions at high temperature ( Tm = ∼72 °C ) , indicating that the protein structures are stabilized by bound peptides , since empty IEk without peptide melts at a much lower temperature and with a broad transition [41] . As expected , when immobilized on ELISA plates , only agonist IEk-MCC induced interleukin 2 ( IL2 ) production by primed AD10 T cells ( Figure 1D ) . Moreover , incubation of the endogenous peptide-IEk proteins ( IEk-ER60 or IEk-HSP70 ) with MCC peptides under conditions reported to load 80% of empty IEks [42] did not endow them with T cell activation capacity ( Figure S2D ) . These data indicate that all three IEk proteins were properly folded with their peptide-binding grooves occupied . Affinity purified IEk proteins were monomers as shown by gel filtration chromatography ( Figure S2A ) , and did not trigger TCR in solution ( unpublished data ) , consistent with previous reports [17–22] . IEk-MCC proteins were never subjected to freezing and thawing , and were always further purified by gel filtration immediately before use . When anchored on the lipid bilayer through HisTag-Ni-NTA binding , IEk proteins moved freely with a diffusion rate only slightly slower than that of the lipids ( Figure S1B ) . The density of IEk on the bilayer was measured at about 2 , 900 per μm2 . GFP-HisTag and ICAM-1-HisTag were also expressed in insect cells for use as a control and in enhancement of T cell adhesion , respectively . Thus , we had in hand a lipid bilayer-based artificial APC system , consisting of a lipid bilayer and proteins with the requisite qualities for study of T cell/APC interactions . To test the ability of agonist pMHCs alone on a fluid surface to trigger TCR , IEk-MCC was mixed with GFP-HisTag at varying ratios prior to anchoring on the lipid bilayer while maintaining a constant concentration of total protein . ICAM-1-HisTag was added at a concentration equal to one-tenth of the total protein concentration . After contacting lipid bilayers with bound IEk-MCC , primed AD10 T cells demonstrated rapid increases in intracellular calcium , as indicated by an increase in the 340 nm/380 nm ratio ( Figure 2A and Video S1 ) . Both the percentage of responding T cells and the amplitude of the increase in intracellular calcium were clearly dependent on the dose of agonist IEk-MCC ( Figure 2B and 2C ) . A 340 nm/380 nm ratio two times above background was observed in T cells interacting with a bilayer anchored with IEk at a GFP:IEk-MCC ratio of up to 100 , 000:1 , whereas no increase in 340 nm/380 nm ratio was observed in T cells interacting with bilayers anchored with only GFP ( Figure 2A and 2C , and Video S2 ) . On the basis of the measured size of primed mouse CD4+ T cells and the density of IEk on the lipid bilayer , we calculated that , on average , three agonist IEk-MCC monomers were sufficient to trigger a low-level calcium response in about 5% of T cells . The same dose-dependent response was observed when diluted IEk-MCC was anchored on the bilayer alone ( Figure 2B ) . To test whether TCR triggering is influenced by the fluidity of the pMHC anchoring surface , we immobilized IEk molecules on a fixed plastic surface . IEk molecules biotinylated on the AviTag sequences were diluted and immobilized on 96-well plates precoated with streptavidin for 18 h at 37 °C for thorough binding . Bio-IEk-MCC–coated wells stimulated AD10 T cells to produce IL-2 in a dose-dependent manner ( Figure 3A ) . Approximately 5% of the T cells were activated by a surface coated with bio-IEk-MCC diluted to a concentration of 1 . 9 × 10−6 μg/ml . Assuming that all bio-IEk-MCC molecules bound to the surface of the well , and using 78 . 5 μm2 as the T cell contact area based on the measured diameter of primed T cells , we calculated that 0 . 83 bio-IEk-MCC molecules per cell , or 83 bio-IEk-MCC per 100 cells , were sufficient for T cell activation . In contrast , surfaces coated with bio-IEk-ER60 , bio-IEk-HSP70 , or bio-IEk-99A did not activate AD10 T cells even at the highest coating concentration , although they bound comparably to streptavidin-coated wells ( Figure 3B ) . IEk-99A contains a covalently linked MCC peptide with a single K>A mutation at the p5 TCR-interacting position that renders it null for AD10 and 5C . C7 T cells [43 , 44] . Thus , T cell activation is induced by very low numbers of agonist , but not endogenous or null , pMHC when presented to T cells on either a fluid or a fixed surface . The conventional notion of TCR crosslinking is by two agonist pMHCs . Because of the very low number of agonist IEk needed to trigger TCR and the highly fluid nature of the lipid bilayer , it was unlikely that agonist–agonist IEk dimers were present on the lipid bilayer . It has been reported , however , that endogenous pMHCs dramatically enhance TCR triggering by agonist pMHCs by formation of “pseudodimers” [28 , 29] . To determine the role of endogenous pMHC in TCR triggering , we replaced GFP-HisTag with IEk-ER60 or IEk-HSP70 . AD10 T cells responded similarly to IEk-MCC in the context of IEk-ER60 or IEk-HSP70 , as well as to IEk-MCC in the context of GFP ( Figure 2A–2C and Videos S1 , S3 , and S4 ) , both in terms of the percentage of responding T cells and the level of calcium increase . Similar results were found using primed 5C . C7 T cells ( unpublished data ) . In addition , we did not observe any difference between IEk-MCC diluted with IEk-HSP70 or with IEk-ER60 in triggered T cell calcium flux patterns ( Figure 2D ) . This is in contrast to a previous report in which only IEk-ER60 , but not IEk-HSP70 , worked synergistically with IEk-MCC , reportedly through pseudodimer formation [29] . To facilitate the formation of pseudodimers between IEk-MCC and endogenous IEks , we immobilized bio-IEk-MCC of varying dilutions onto streptavidin-coated 96-well plates . This was followed by incubation with coating buffer , bio-IEk-ER60 , bio-IEk-HSP70 , or bio-IEk-99A at an experimentally confirmed saturating concentration to occupy remaining free biotin binding sites . We reasoned that binding of bio-IEk-MCC and endogenous pMHC on the same tetravalent streptavidin molecule under these conditions would form stably linked pseudodimers . The distance between pMHCs , but not their orientation , is reportedly important for TCR crosslinking [45] , with comparable crosslinking demonstrated for linkers ranging from a direct disulfide bond ( 0 . 2 nm ) to a peptide crosslinker of 7 nm in length [45] . The distance between two biotin binding sites ( ∼3 nm ) on streptavidin is well within this range [46] . Indeed , streptavidin bound with two or more IEk-MCC is capable of activating T cells in solution [22] . Consistent with our observed calcium responses to IEk on lipid bilayers , saturating the surface with endogenous bio-IEk-ER60 , bio-IEk-HSP70 , or bio-IEk-99A did not enhance the ability of bio-IEk-MCC to induce IL2 production by AD10 T cells ( Figure 3C ) or 5C . C7 cells ( Figure 3E ) . These results were confirmed by experiments showing that activation of AD10 T cells ( Figure 3D ) or 5C . C7 T cells ( Figure 3F ) by bio-IEk-MCC directly bound to ELISA plates was not enhanced by subsequent saturation binding by IEk-ER60 , IEk-HSP70 , IEk-99A , or IAk-CA . IAk coexists with IEk on the same APCs in vivo . IAk with covalently linked conalbumin peptide ( IAk-CA ) activates T cells bearing D10 TCRs ( Figure S3 ) . On the ELISA plate , the density of bio-IEk-MCC remained the same with or without subsequent saturation binding by other proteins ( Figure S4 ) . The difference in shape and range of the response curves to IEk on streptavidin versus ELISA plates is likely a result of their different surface properties ( see Text S1 ) . Therefore , on fluid lipid bilayers or on fixed surfaces , T cell activation induced by agonist IEk-MCC was not enhanced by endogenous IEk , null IEk , or another MHC class II molecule , IAk . To test the role of endogenous pMHCs in TCR triggering under more physiologic conditions , we skewed the makeup of endogenous peptides on real APCs . We wished to determine whether the ability of APCs to activate T cells could be altered by replacing a major proportion of the original diverse endogenous peptide population with a peptide having putative enhancing or nonenhancing effects on TCR triggering . Murine B-cell lymphoma CH27 cells expressing IEk were incubated with biotinylated β2m peptides ( β2m-bio ) , ER60 peptides ( ER60-bio ) , or scrambled ER60 peptides ( ER60scrbl-bio ) at high concentrations for 20 h . To determine the degree of IEk occupancy by the biotinylated peptides , pulsed cells were stained with streptavidin-Cy5 , and the fluorescence intensity was compared with CH27 cells stained with biotinylated anti-IEk monoclonal antibody , 14-4-4s ( 14-4-4s-bio ) , followed by streptavidin-Cy5 . The peptides were biotinylated at the N-terminus , so each peptide bound to one streptavidin-Cy5 . The number of streptavidin-Cy5 each 14-4-4s-bio antibody could bind was determined using a gel filtration chromatography-based assay ( Figure S5 ) , and the result was used to correct for differences in fluorescence intensity . Using this assay , we determined that β2m-bio , ER60-bio , or ER60scrbl-bio peptides occupied about 59% , 17% , or 3% , respectively , of IEk molecules on the surface of CH27 cells ( Figure S6 ) . The peptide binding was IEk-specific; only a very low level of binding was observed on the mutant murine B-cell lymphoma , M12 . C3 , which is deficient for MHC class II ( Figure S6 ) . The CH27 cells were then briefly incubated with FITC-labeled MCC peptides ( MCC-FITC ) before adding to a T cell stimulation assay . MCC-FITC pulsing led to comparable FITC levels on CH27 cells prepulsed with β2m-bio , ER60-bio , or ER60scrbl-bio ( Figure 4A ) , while maintaining relatively consistent IEk occupancy by biotinylated endogenous peptides ( unpublished data ) . In a previous report , only ER60 , but not β2m , enhanced T cell activation , presumably by forming pseudodimers with agonist pMHC [29] . If this were the case , then skewing the endogenous peptide population on the CH27 surface towards ER60 or β2m should significantly enhance or reduce , respectively , the chance of pseudodimer formation , and lead to different degrees of T cell activation upon stimulation by comparable levels of agonist MCC peptides . In contrast , in our hands , T cells stimulated with CH27 cells prepulsed with either ER60-bio or β2m-bio showed very similar IL2 production to cells prepulsed with the control ER60scrbl-bio peptide ( Figure 4B ) , a peptide which did not significantly alter the original endogenous peptide makeup . Therefore , on real APCs , agonist pMHCs appear to operate independently of endogenous pMHCs in T cell stimulation . Previous work [22] and our own experimental data argue strongly that monomeric agonist pMHC cannot trigger TCR in solution . To determine what confers the ability of a very small number of agonist pMHCs , anchored on a surface , to trigger TCR , we examined whether active adhesion was required for triggering by our artificial APCs . T cells spontaneously adhered to lipid bilayers containing NTA-Ni ( unpublished data ) , probably due to surface charge interactions . The inability of this bilayer to become a nonadhesive surface made it unsuitable for studies of the role of adhesion in TCR triggering . Therefore , in order to have in hand a lipid bilayer system with controllable adhesion properties , we established a biotin-streptavidin–based artificial APC system ( Figure S7 ) . Streptavidin molecules specifically bind to POPC bilayers containing 5 mol% DOPE-biotin and diffuse freely ( Figure S1C ) . IEk and ICAM-1 molecules were biotinylated at their AviTag sequences and mixed with streptavidin at ratios suitable for forming complexes with desired valencies . Monovalent ( IEk-MCC ) -streptavidin and divalent ( ICAM-1 ) 2-streptavidin complexes were further purified by gel filtration . T cell adhesion to this bilayer was strongly dependent on the presence of ( ICAM-1 ) 2-streptavidin on the bilayer ( Videos S5 and S6 ) . Bilayers anchored only with ( IEk-MCC ) -streptavidin failed to stimulate AD10 T cell calcium flux , even at relatively high ligand densities ( 1 , 400/mm2 ) ( Figure 5A and Video S7 ) . In contrast , in the presence of ICAM-1 , bilayers anchored with ( IEk-MCC ) -streptavidin induced strong calcium flux ( Figure 5B and Video S8 ) , suggesting that adhesion is required for TCR triggering by agonist pMHC on the lipid bilayer . In addition to adhesion , another important aspect of the dynamic 2D interface formed by two highly mobile cells , such as the T cell and its APC , is the impact of cytoskeletal movement , which may directly apply force upon pMHC-TCR binding . To test the role of the actin cytoskeleton in TCR triggering by agonist pMHC , AD10 T cells were treated with the actin depolymerizing agent , cytochalasin D , before interaction with a layer of CH27 cells pulsed with MCC peptides on glass . While untreated AD10 T cells responded with pronounced and sustained increases in intracellular calcium ( Figure 5C and Video S9 ) , AD10 T cells pretreated with cytochalasin D showed no response ( Figure 5D and Video S10 ) , although in both cases , there was contact between T cells and APCs ( Figure S8 ) . Similar results were observed when T cells were stimulated with IEk-MCC anchored on fluid lipid bilayers or plastic surfaces ( Figure S9 ) . Therefore , cytochalasin D treatment completely blunted T cell calcium responses to real or artificial APCs . To test whether actin cytoskeletal rearrangement is required for sustaining calcium flux that has already been initiated by real APCs , cytochalasin D was introduced after the calcium flux of untreated T cells reached peak levels . In agreement with a previous report [47] , cytochalasin D quickly suppressed calcium flux ( Figure 5E ) . Cytochalasin D treatment per se did not block events downstream of TCR triggering leading to T cell calcium flux , since calcium flux induced by anti-TCR antibody crosslinking was unaffected ( Videos S11–S13 ) . Cytochalasin D treatment also significantly reduced T cell mobility and morphological changes associated with interaction with CH27 cells ( unpublished data ) . Intact actin cytoskeletal function is therefore required for TCR triggering by both artificial APCs and real APCs .
We demonstrate , using artificial APCs , that surface anchoring , adhesion , and actin cytoskeleton enable monomeric agonist pMHCs to trigger TCR with remarkable potency . T cells responded to an average of three and of 0 . 83 agonist pMHCs anchored on lipid bilayers and fixed plastic surfaces , respectively . Due to the nonaverage Poisson distribution of the ligands on the surfaces , the responding T cells might have contacted as many as seven and three agonist pMHCs , respectively ( see Text S2 ) . Although it is possible that T cells may have responded to a larger number of agonist pMHCs on lipid bilayers due to their diffusion and accumulation underneath T cells over time , the general agreement between the required numbers of ligands on lipid bilayers and plastic surfaces suggests that TCR can be triggered by a very low number of agonist pMHCs anchored on fluid or fixed surfaces . We believe that an alternative interpretation of our data , i . e . , that TCR was triggered by agonist pMHC dimers , is unlikely for the following reasons . First , we took great caution in generating and purifying IEk-MCC proteins to minimize the existence of dimers in our protein preparation . The very low number of surface-anchored agonist pMHCs required for TCR triggering ( <10 per cell ) ( Figures 2A–2C , 3A , and Text S2 ) makes it highly unlikely that TCR was triggered by dimer contamination in the protein preparation , which certainly did not contain >10% dimers ( Figure S2A ) . Second , to exclude the possibility of “de facto” dimer formation by immobilization of two agonist pMHCs within close proximity by chance , we developed the lipid bilayer-based artificial APC system and went to great lengths to ensure the homogeneity and fluidity of the bilayer . Third , although dimers might , theoretically , form spontaneously between monomeric pMHCs diffusing freely on the bilayer , the necessary intrinsic affinity between MHC molecules is not supported by experimental evidence . Despite the availability of such highly sensitive techniques as surface plasmon resonance , no quantitative measurement of affinity between MHC molecules has been documented , suggesting that the affinity between MHC molecules is extremely low , if there is any . This is consistent with the fact that MHC-MHC dimers have not been observed in the vast majority of MHC crystallography studies [7 , 9–11 , 48–52] . The possibility of such a scenario is further reduced by the presence of a large number of “diluting” endogenous or null pMHCs . The lack of affinity between MHC molecules may not be relevant , if TCRs exist as dimers on the T cell and “recruit” two pMHC monomers sequentially . Whether TCRs are present as dimers on the T cell surface , however , is highly controversial [53–58] . Even if TCRs were dimers , the fast dissociation rate of agonist pMHC-TCR binding under force ( discussed below ) would make it very difficult to successfully recruit the second agonist pMHC molecule , especially when the number of agonist pMHCs is low . Finally , the shape of the T cell dose response curve further supports the sufficiency of monomeric pMHC for triggering . T cells responded to increasing numbers of IEk-MCC on the bilayer or on fixed surfaces in a sublinear fashion , as opposed to a second order dependence ( y = ax2 ) that would have indicated a requirement for IEk-MCC dimers ( Figure 6 ) . The sublinear response is consistent with the wide dynamic range observed in T cell responses to IEk-MCC on both lipid bilayers and plastic surfaces , and may indicate a progressive suppression of T cell sensitivity to increasing doses of agonist pMHCs . The physiological significance of this phenomenon remains to be investigated , but similar responses have been observed in other biological systems , such as the response of rod cells to photons triggering the photoreceptor , rhodopsin [59] . In summary , our data indicate that individual agonist pMHC monomers are sufficient to trigger TCR . In a recent study [29] , help from endogenous pMHCs was used to explain the remarkable T cell sensitivity to agonist pMHCs on real APCs . A pseudodimer model was proposed in which an endogenous pMHC engaging TCR could be bridged by CD4 to form a “pseudodimer” with an agonist pMHC engaging another TCR [29] . This model of TCR triggering was based upon experimental data showing that the very low TCR triggering capacity of agonist pMHC alone ( requiring a density of more than 1 , 000 per cell on lipid bilayers ) [29 , 60] was dramatically enhanced ( more than 100-fold ) by the presence of endogenous pMHCs . Our observation that one to ten agonist IEk-MCC alone triggered not only TCR-induced calcium flux , but also IL2 production , calls into question the purported requirement for help by endogenous pMHC . Indeed , using mouse T cells with a transgenic TCR of the same specificity , endogenous IEk-ER60 and null IEk-99A , which were shown to enhance TCR triggering by IEk-MCC in the previous report , did not , in our hands , enhance TCR triggering by IEk-MCC on fluid lipid bilayers or fixed plastic surfaces . As evidenced by our analyses using circular dichroism and peptide pulsing experiments , the lack of help by IEk-ER60 could not be attributed to incorrect protein folding or lack of peptide in the binding groove ( Figure S2B–S2D ) . IEk-99A differs from IEk-MCC by only one residue , and the correctness of its tertiary structure is supported by the fact that tetrameric IEk-99A weakly binds T cells from AND TCR transgenic mice [35] and acts as a weak agonist for these T cells ( Figure S10 and S . M . Hedrick , personal communication ) . The lack of help for TCR triggering by endogenous pMHC was further supported by our observation that biasing the endogenous peptide population on real APCs with ER60 or β2m peptides did not affect T cell activation by MCC peptides . This discrepancy may be partially explained by the different methodologies used in these studies . We consistently used MHC molecules with covalently linked peptides expressed in insect cells . In the previous report , both MHC with covalently linked peptides and empty MHC loaded with synthetic peptides were used , sometimes in the same experiment . The relative MHC occupancies by the different peptides were not determined . The differing results of T cell activation by real cells pulsed with peptides may be due to the fact that two totally different cell types were used in the two studies: previously , CHO cells expressing empty GPI-anchored IEk versus , in our study , CH27 B cells expressing IEk , as well as the appropriate adhesion and costimulatory molecules of real APCs . Again , in the previous study , the relative level of different peptides on the surface of CHO cells was not determined . Theoretically , some aspects of the pseudodimer model of TCR remain to be reconciled with what has been learned regarding TCR triggering and pMHC-TCR interaction . The hypothesis that CD4 bridges the pseudodimer is not well supported by experimental data . Coreceptors bind MHC with very limited affinity and stability and with a fast off-rate [61] . In addition , CD4 has been shown to associate with TCR/CD3 via Lck and ZAP-70 in response to CD3 stimulation [62] , but not on resting T cells . Another issue is that the role of endogenous pMHC is not clearly explained in the pseudodimer model . Intuitively , it seems that recruitment of free TCR-CD4 to an agonist pMHC-TCR binding pair would be easier than recruitment of the TCR-CD4 complex bound to endogenous pMHC anchored on an opposing membrane . The resulting complex should function equally well as a pseudodimer . Finally , using TAP-deficient cells , previous reports demonstrated that endogenous pMHCs had a negligible effect on the activation by agonist pMHCs of CD8+ T cells , which express the same TCR/CD3 complex [24 , 63] . What is it , then , that transforms a nonfunctional soluble monomeric agonist pMHC in solution into a powerful TCR triggering unit once it is attached to a surface ? Recent data suggest that the mechanism of TCR triggering must include features unique to 2D binding . These include the need for two opposing membranes to reach the confinement length for association of membrane-anchored ligands and receptors , especially ones with small dimensions such as TCR and pMHC [31 , 32 , 64] . What has been omitted , however , is that a bound ligand–receptor pair , including pMHC-TCR , is constantly stressed and may be ruptured by forces from the active cytoskeletal rearrangement that supports the dynamic engagement between T cells and APCs . Recent in vivo studies using two-photon microscopy have characterized the interaction between T cells and antigen-loaded APCs in lymph nodes during the first 2 h as serial and dynamic , with T cells engaging and disengaging APCs at a speed of 5 . 4 μm/min [65 , 66] . T cells move with a velocity of 2 . 6 μm/min even at later stages of dynamic T cell/APC clustering . In this context , a single specific interaction between a TCR and an agonist pMHC may not be able to maintain the small , close membrane–membrane contact zone and keep tyrosine kinases and phosphatases segregated , as proposed in the kinetic-segregation model [30–32] . We demonstrate , in this study , that TCR triggering requires not only T cell adhesion to a surface , but also a functional actin cytoskeleton , consistent with the possibility that detaching forces are important in TCR triggering . Here , taking into consideration the dynamic aspect of the 2D interaction between T cells and APCs , especially its likely impact on the dissociation of pMHC-TCR , we propose the receptor deformation model of TCR triggering ( Figure 7 ) . In this model , when a T cell encounters and scans an APC , pMHC and TCR interact when and where a sufficiently close membrane–membrane contact is formed . pMHC-TCR interaction per se , however , does not trigger TCR . A signal is initiated when the binding pair is pulled by a detaching force originating from cytoskeletal rearrangement , a force originally intended to detach the membrane–membrane contact for T cell movement . The pulling force induces a conformational change of the αβ TCR , which is then transferred to the intracellular domains of the CD3 complex through interactions between their extracellular domains and transmembrane domains [67] . Alternatively , it may directly cause changes in the relative positioning and orientation of the components of the TCR/CD3 complex . This pulling force may also induce conformational changes of the coreceptor binding domains of MHC molecules , and lead to better recruitment or conformational change of the coreceptors . The signal is initiated by increased access to and phosphorylation of ITAM domains of the CD3 complex by the coreceptor-associated tyrosine kinase , Lck . Therefore , the critical factor that determines whether a particular pMHC-TCR binding can lead to signaling is whether the binding has sufficient mechanical strength and appropriate kinetics to deliver an external force to TCR or pMHC to induce a conformational change , rather than the affinity , kinetics , or conformational change under zero force , as proposed in previous models . The receptor deformation model provides a solution to the problem of how the binding between TCRs and pMHCs , both of which have hugely variable binding interfaces , leads to a uniform signal initiating conformational change of the TCR/CD3 complex . It also explains how T cells can detect an extremely low number of agonist pMHCs , alone , anchored on surfaces , as we observed in this study . Although under zero force , the half-life of binding ( t1/2 ) of many agonist pMHC-TCR pairs has been reported to be more than 10 s , and as long as 50 s for strong agonists [68] , it could be much shorter under a pulling force , because the dissociate rate ( koff ) increases exponentially with force [69] . Together , with the locomotion and dynamic morphological changes of the T cells , and the active lateral movement of TCR and pMHC on the fluid membranes [70] , a fast dissociation could allow rapid triggering of multiple neighboring TCRs by a single agonist pMHC in a short period of time , leading to efficient temporal and spatial accumulation and integration of multiple signals , as previously proposed by the serial engagement model [71] . Furthermore , the receptor deformation model offers the “rupture force” of pMHC-TCR binding ( the force needed to rupture the binding ) as the mechanism by which T cells distinguish structurally similar agonist and endogenous pMHCs . This is supported by the excellent correlation found between rupture force and the zero force koff [72] , which is correlated with TCR triggering by pMHC [68] . It is intriguing to speculate that weak rupture forces between TCR and endogenous pMHC may support minor changes of the TCR/CD3 complex and generate survival signals . Finally , this model is consistent with experimental evidence that cognate pMHC-TCR interaction induces a conformational change in the CD3 complex [73 , 74] . Also , force-induced conformational change has been documented in other ligand-receptor systems , such as the conformational change of LFA-1 α domain [75] and Escherichia coli adhesin FimH [76] in response to shear forces . The receptor deformation model is very different from a previously proposed raft-TCR collision model that also involves force , in which the movement of lipid rafts on T cells driven by actin or shear stress promotes collision of raft-associated Lck and the TCR/CD3 complex [77] . This model works through increased kinase-TCR proximity , rather than conformational change . In the raft-TCR collision model , the force is applied laterally on the lipid raft and does not impact the binding kinetics of pMHC-TCR interaction . Our model is also distinct from the permissive geometry model , which proposes that TCRs are present as oligomeric clusters on T cells , and that binding of multimeric agonist pMHCs with a certain permissive geometry triggers TCRs through changes in the relative positioning of individual TCRs within the cluster [57 , 78] . The key differences are that changes in TCR geometry are induced solely by pMHC-TCR interaction , without the requirement of external mechanical force , and that multimeric pMHCs are required for triggering . Notwithstanding its uniqueness and advantages , the receptor deformation model needs to be reconciled with the well-documented observation that soluble multimeric agonist pMHCs trigger TCR [22 , 79] . In our opinion , the fact that T cells can be activated via a receptor crosslinking mechanism used by B cells is not surprising , given their similarity in evolution , development , antigen receptor structure , and intracellular signaling pathways . Our work does not exclude the possibility that when agonist pMHCs are present at extremely high levels on APCs , TCR crosslinking is a functional mechanism of TCR triggering . In this situation , the chance of two neighboring MHC molecules , both presenting agonist peptides , may be reasonably high . This , and the fact that a high proportion of MHC molecules on the APC surface is immobile [26 , 27] , make it plausible that two monomeric agonist pMHCs might be stably localized in close enough proximity to act as functional dimers capable of crosslinking TCRs . Given the rarity of agonist pMHCs on APCs under physiological conditions , however , a mechanism that does not rely on TCR crosslinking , such as receptor deformation , is likely to be the main working mechanism of TCR triggering . In conclusion , we provide strong evidence that extremely small numbers ( <10 per cell ) of surface-anchored agonist pMHCs trigger TCR , independent of endogenous pMHCs , but dependent upon adhesion and intact cytoskeletal function . We propose the receptor deformation model , in which cytoskeletal rearrangement delivers the driving force for TCR triggering . Given its merits in providing a straightforward mechanism of TCR triggering and in explaining both the high sensitivity and the high specificity of agonist pMHC detection by T cells , we believe that this model warrants further investigation .
B10 . BR H-2k mice and 5C . C7 mice were purchased from Jackson Laboratories and Taconic , respectively . AD10 and AND TCR transgenic mice were provided by J . W . Kappler ( National Jewish Medical and Research Center ) . T cell blasts were generated by a mixed culture of splenocytes from TCR transgenic mice and B10 . BR mice in Click's medium supplemented with 10% fetal calf serum ( FCS ) , 2 mM l-glutamine , 50 μM β-mercaptoethanol , penicillin/streptomycin , 1 mM sodium pyruvate , 0 . 1 mM nonessential amino acids , and 50 μg/ml pigeon cytochrome c . T cell blasts were used on day 7 to day 9 poststimulation . POPC , DOPE-biotin , DOPE-NBD , and DOGS-NTA were purchased from Avanti Polar Lipids; 14-4-4s mAb was generated from a hybridoma kindly provided by J . W . Kappler . Anti-ICAM-1 monoclonal antibody , YN1 . 7 . 4 , was produced from a hybridoma obtained from ATCC . Anti-leucine zipper antibody , 2H11 , was produced from a hybridoma from E . L . Reinherz ( Dana-Farber Cancer Institute ) . Streptavidin was from Sigma . APC-labeled anti-mouse IL2 antibody , JES6-5H4 , and APC-labeled anti-mouse IL4 antibody , 11H11 , were from BD Biosciences . The D10 . IL2 T cell line , CH27 , and M12 . C3 B cell lymphoma cells were provided by A . Kupfer ( Johns Hopkins University ) , J . Monroe ( University of Pennsylvania ) , and S . Ostrand-Rosenberg ( University of Maryland ) , respectively . ER60-bio , β2m-bio , ER60scrbl-bio ( APPNIKYFLSFGTK ) , and MCC-FITC peptides were synthesized by Genemed Synthesis . Baculovirus transfer vector , pBlueBac4 . 5/V5-His , Sf9 , and Hi5 insect cells were purchased from Invitrogen . All proteins were expressed in secreted form by infecting Hi5 insect cells with baculovirus . Baculovirus transfer vector for ICAM-1-AviTag-HisTag was constructed with pBlueBac4 . 5/V5-His and murine ICAM-1 cDNA amplified from mouse spleen mRNA . Transfer vectors for IEk-MCC , IEk-HSP70 , IEk-ER60 , IEk-99A , and IAk-CA with AviTag and HisTag were constructed based on a transfer vector for IEk-MCC-AviTag ( a gift from J . W . Kappler ) . cDNA encoding IAk-CA stabilized with a leucine zipper was provided by E . L . Reinherz . ICAM-1 , IEk , and IAk proteins were affinity purified with columns conjugated with NTA-Ni , 14-4-4s antibody , and 2H11 antibody , respectively . Each protein was then further purified by gel filtration with a Superdex 200 10/300 GL column ( Amersham Biosciences ) before storage . IEk-MCC was kept at 4 °C for less than 10 d and was never frozen and thawed . An additional round of gel filtration was always performed immediately before use of the proteins in experiments . IEk-MCC and ICAM-1 were biotinylated with BIRA enzyme ( Avidity ) and purified by gel filtration . The biotinylation rate was about 50% , as measured by an ELISA-based assay . To generate ( IEk-MCC ) -SA , bio-IEk-MCC and streptavidin were mixed at a 1:5 molar ratio , and monovalent complexes were purified by two successive rounds of gel filtration with a Superdex 200 10/300 GL column . To generate ( ICAM-1 ) 2-SA , bio-ICAM-1 and streptavidin were mixed at a 1:1 molar ratio , and bivalent complexes were purified by gel filtration . Purified complexes were used immediately without storage . Liposomes were prepared by sonication using POPC mixed with 0 . 2 mol% DOPE-NBD and 5 mol% DOGS-NTA or DOPE-biotin in HEPES-Na buffer ( 150 mM NaCl , 5 mM HEPES [pH 7 . 4] with 0 . 02% sodium azide ) . The liposome solution was then spun at 40 , 000 rpm for 2 h , and small unilamellar vesicles ( SUV ) were isolated by taking the top four-fifths of the solution . Supported planer lipid bilayers were prepared by liposome fusion for 10 min at 4 °C on glass cover slips that had been extensively cleaned with Contrad70 detergent ( Decon Laboratories ) and chromic sulfuric acid solution ( Sigma ) . Half of the lipid bilayer was collapsed by exposure to air before transferring to 25 °C for 40 min for expansion . Lipid bilayers were then transferred under HEPES-Na buffer into FCS2 flow chambers ( Bioptechs ) for imaging . For bilayer containing DOGS-NTA , HEPES-Na buffer containing 0 . 25 mM NiSO4 was introduced into the chamber for 10 min to charge DOGS-NTA with Ni2+ . Lipid bilayers were blocked with blocking buffer ( 5 mg/ml BSA in HEPES-Na buffer [pH 7 . 4] ) for 10 min . For lipid bilayers with DOGS-NTA , 10 μg/ml IEk and 1 μg/ml ICAM-1 in blocking buffer were introduced into the chamber and incubated for 30 min at 25 °C . The bilayer was then washed with imaging buffer ( 150 mM NaCl , 5 mM KCl , 2 mg/ml glucose , 10 mM HEPES [pH 7 . 4] , 33 mg/ml BSA ) containing 1 mM CaCl2 and 1 mM MgCl2 for 3 min under constant buffer flow . For bilayers with DOPE-biotin , the blocked bilayer was incubated with 1 μg/ml ( IEk-MCC ) -SA first , followed by 5 μg/ml ( ICAM-1 ) 2-SA in imaging buffer . The bilayer was then washed with imaging buffer containing 0 . 25 mM MnCl2 and 0 . 25 mM CaCl2 . For POPC bilayers with 5 mol% DOGS-NTA , IEk-MCC was labeled with FITC . For POPC bilayers with 5 mol% DOPE-biotin , biotinylated IEk-MCC was labeled with FITC and was used to form the ( FITC-IEk-MCC ) -SA complex . After binding of the FITC-labeled proteins , the bilayers were lysed with 200 μl of DPBS ( pH 7 . 4 ) containing 1% Triton-X100 and 0 . 5 mg/ml BSA . The intensity of FITC was read using a high numerical aperture ( 0 . 5 ) 10× air objective ( Zeiss ) and recorded with a 12-bit cooled CCD camera . POPC bilayer without DOGS-NTA or DOPE-bio was used as a blank control . After subtracting the FITC intensity of the blank , the FITC intensity was translated into protein concentration based on a standard curve generated using FITC-labeled proteins of known concentration . The ligand densities were calculated based on the amount of protein bound and the area of the lipid bilayer . For calcium imaging , 5 × 106 TCR transgenic CD4+ T cells were pulsed with 5 μM fura-2-AM ( Invitrogen ) for 30 min at room temperature and washed twice with imaging buffer before introduction into the flow chamber . The 510-nm emissions excited by 380 nm and 340 nm were captured at 5-s intervals for 30 min using a 40× oil Plan-Neofluar objective on an Axioplan2 Microscope ( Zeiss ) at 25 °C . Data were collected and analyzed with SlideBook software ( Intelligent Imaging Innovations ) . Nomarski DIC bright-field imaging with a 60× water Achroplan objective was used to capture cell adhesion on the lipid bilayer and to measure the T cell diameter . For FRAP , a pattern of 4-μm radius was photobleached with a 3 mW nitrogen laser ( Stanford Research Systems ) , and the recovery was captured using an attenuated xenon light source ( Sutter ) . Bio-IEk-MCC in PBS pH 8 . 0 was coated on Immulon2 U-bottom ELISA plates ( Thermo Electron ) or SigmaScreen streptavidin-coated plates ( Sigma; binding capacity ≥6 pmol biotin/well ) for 18 h at 4 °C or 37 °C , respectively . For experiments requiring saturation of the plate by another coating protein , 10 μg/ml nonbiotinylated or biotinylated proteins were subsequently added to ELISA or streptavidin plates in a 50 or 75 μl volume , respectively . The ELISA or streptavidin plate was then incubated at room temperature for 4 h or 1 h , respectively . Plates were washed , and 2 . 5 × 105 T cells were added in complete Click's medium containing 20 μg/ml brefeldin A . After 7 h incubation , cells were harvested , fixed with 3% formaldehyde in PBS , permeabilized with PBS buffer containing 1% BSA and 0 . 1% saponin , and stained with anti-IL2 mAb for flow cytometric analysis . CH27 cells at 5 × 105/ml were incubated with 1 mM β2m-bio , ER60-bio , or Er60srcbl-bio peptides in Click's medium with 10 mM HEPES for 20 h at 37 °C . After washing three times with Click's medium , cells were incubated for 30 min at 37 °C with MCC-FITC peptides at different dilutions . After washing , half of the cells pulsed at each dilution were then stained with streptavidin-Cy5 , and the FITC and Cy5 intensities were measured by flow cytometry . The other half of the cells were incubated with AD10 T cells at a 1:1 ratio for assay of IL2 production . | Using the T cell receptor ( TCR ) as a sensor , T cells of the immune system constantly migrate in lymphoid organs and probe the surface of antigen-presenting cells ( APCs ) for foreign antigens , a sign of pathogen infection . Antigen binding by TCRs leads to T cell activation and subsequent immune response to combat the pathogens . Interestingly , although T cells respond well to antigens on APCs , they do not recognize the same antigens in solution . What is it that makes antigens on APCs recognizable ? To address this , we used lipid bilayers and plastic surfaces to construct artificial APCs with defined antigen number , composition , and configuration . We found that T cells respond to very few individual foreign antigens on artificial APCs , and contrary to some current opinion , formation of antigen clusters on APCs is not required for antigen recognition by T cells . TCR triggering , however , requires T cell adhesion to the APC surface and then occurs only if the T cells are able to move . We propose that at the dynamic T cell–APC interface , antigen on APCs activates T cells by applying force to the TCR and deforming its structure , which cannot be achieved by soluble antigens due to their lack of anchorage . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biotechnology",
"biochemistry",
"cell",
"biology",
"immunology",
"biophysics"
] | 2008 | Surface-Anchored Monomeric Agonist pMHCs Alone Trigger TCR with High Sensitivity |
The mechanisms by which the sensory environment influences metabolic homeostasis remains poorly understood . In this report , we show that oxygen , a potent environmental signal , is an important regulator of whole body lipid metabolism . C . elegans oxygen-sensing neurons reciprocally regulate peripheral lipid metabolism under normoxia in the following way: under high oxygen and food absence , URX sensory neurons are activated , and stimulate fat loss in the intestine , the major metabolic organ for C . elegans . Under lower oxygen conditions or when food is present , the BAG sensory neurons respond by repressing the resting properties of the URX neurons . A genetic screen to identify modulators of this effect led to the identification of a BAG-neuron-specific neuropeptide called FLP-17 , whose cognate receptor EGL-6 functions in URX neurons . Thus , BAG sensory neurons counterbalance the metabolic effect of tonically active URX neurons via neuropeptide communication . The combined regulatory actions of these neurons serve to precisely tune the rate and extent of fat loss to the availability of food and oxygen , and provides an interesting example of the myriad mechanisms underlying homeostatic control .
The central nervous system plays a critical role in regulating whole body energy balance . In mammals , in addition to the role of the hypothalamus , there is good evidence showing that the sensory nervous system also regulates whole body metabolism [1–4] . However , the identification of discrete sensory modalities and the underlying signaling pathways that regulate metabolism in peripheral tissues has remained a tremendous challenge . As a result , basic questions regarding the molecular and physiological mechanisms underlying the neuronal control of lipid metabolism remain unknown even as rates of obesity , diabetes and their complications soar worldwide . In the nematode Caenorhabditis elegans , many ancient mechanisms of sensory and metabolic regulation have been preserved , thus offering an opportunity to dissect the pathways by which the nervous system regulates lipid metabolism . The C . elegans nervous system is relatively well-defined both genetically and anatomically [5 , 6] , thus genes underlying discrete sensory modalities can be rapidly assessed for roles in whole body metabolism . The intestine is the seat of metabolic control and is the predominant depot for fat uptake , storage and mobilization . Thus , changes in whole body metabolism can be effectively encapsulated by monitoring metabolic readouts in the intestine . C . elegans sensory neurons play pivotal roles in regulating metabolism [7 , 8] . At least 3 discrete sensory inputs: food availability [9–11] , population density [12] and environmental oxygen [13] are relayed from chemosensory neurons to the intestine . These sensory inputs control the duration and magnitude of fat loss , effectively coupling environmental information with body fat metabolism via neuroendocrine hormones . The molecular and neuroendocrine mechanisms by which sensory information is relayed from the nervous system to the intestine has the tremendous potential to shed light on the interaction between genetic mechanisms and environmental conditions in controlling lipid metabolism . Neuronal oxygen sensing is a sensory modality that is richly informative to C . elegans , which grow in environments replete with bacteria , their major food source . Respiring bacteria drop the local ambient concentration of oxygen from 21% ( atmospheric ) to a range between 10–13% . Worms presented with an oxygen gradient from 0–21% choose a range between 10–13% , and avoid higher and lower oxygen concentrations , the better to remain in areas that reflect the balance between sufficient food and oxygen [14 , 15] . Importantly , these responses are distinct from hypoxia-related responses , which occur at 3% oxygen and below , are encoded by the conserved hypoxia genes including HIF-1 , and are distinct from mechanisms of oxygen sensing during normoxia [16] . Oxygen preference within the normoxic range is encoded by two pairs of sensory neurons [17] . The URX neurons detect high ( 21% ) oxygen via a guanylate cyclase called GCY-36 , and initiate an aversive behavioral response [17 , 18] . On the other hand , BAG neurons detect low ( 5–10% ) oxygen via an inhibitory guanylate cyclase called GCY-33 , and also initiate aversion [17 , 19] . Thus , URX and BAG neurons function together to retain animals in food environments that contain an optimal balance of food and oxygen . In a screen for neuronal regulators of fat metabolism , we uncovered a role for the URX neurons in regulating oxygen-dependent fat loss [13] . When fasted , body fat stores are metabolized for the production of energy . A time course of food withdrawal at atmospheric ( 21% ) oxygen shows that the intestinal fat stores of C . elegans are steadily depleted , and by 3 hours young adults lose ~80% of their body fat stores; near-complete fat loss in the intestine occurs by 4 hours [13] . Wild-type animals exposed to 10% oxygen retain approximately twice as much body fat as those exposed to high oxygen . We previously showed that this effect is not regulated by changes in food intake or locomotor rates at high versus low oxygen [13] . Rather , it occurs because of a selective metabolic shift towards fat utilization that is differentially regulated by oxygen availability . In the intestine , this metabolic shift depends upon the conserved triglyceride lipase called ATGL-1 , which is transcriptionally regulated by oxygen . Thus , animals exposed to high oxygen show greater fat loss via ATGL-1 activation , than those exposed to low oxygen for the same duration . This differential metabolic response to oxygen depends on neuronal oxygen sensing via the URX neurons , and functions via a sensor of molecular oxygen , the soluble guanylate cyclase GCY-36 . We found that a gustducin-like Gα protein called GPA-8 functions as a negative regulator of GCY-36 , and serves to limit the tonic , constitutive activity of URX neurons by repressing resting Ca2+ levels . Loss of GPA-8 disinhibits the tonic activity of the URX neurons at 10% oxygen ( when they are normally silent ) , which leads to increased fat loss [13] . Our findings on the role of the URX neurons in regulating oxygen-dependent fat loss led us to contemplate a potential role for the BAG neurons , which reciprocally regulate behaviors associated with sensing low oxygen ( 5–10% ) and more broadly , of the role of normoxic oxygen in regulating lipid metabolism [17 , 19] . The URX and BAG neurons are activated at high and low oxygen , respectively [14 , 15 , 17 , 19–21] . However , these neurons are known to be tonic sensors of oxygen , suggesting that they would need to be kept in the 'off state' by active repression . Neuronal mechanisms underlying repression of tonically active neurons are not well-defined . In this study , we report that BAG neurons play an important role in regulating oxygen-dependent fat loss . We find that BAG neurons relay sensory information about low oxygen to the URX neurons and inhibit them to maintain low tonic activity . The physiological consequence of this neuronal signaling is to reduce the rate and extent of fat loss in the intestine during low oxygen . Communication from BAG neurons to URX neurons occurs via the FLP-17 neuropeptide from the BAG neurons and its cognate receptor , the G protein coupled receptor EGL-6 , which we show functions in the URX neurons . Thus , we find that BAG neurons play a critical role in modulating body fat homeostasis by controlling the resting properties of URX neurons . Our work points to a model in which neurons with opposing sensory roles establish a mutually reinforcing circuit via neuropeptide signaling , and function together to regulate metabolic homeostasis in the intestine .
Because of our previously-described role for the URX-neuron-specific soluble guanylate cyclase GCY-36 in regulating oxygen-dependent fat loss [13] , we examined all available null mutants of the soluble guanylate cyclase ( sGC ) family for changes in body fat ( Figs 1A , S1A and S1B ) . Relative to wild-type animals gcy-33 mutants showed a ~40–50% reduction in body fat stores , whereas other sGC loss-of-function mutants did not show an appreciable difference . Biochemical extraction of triglycerides from whole animals in wild-type and gcy-33 mutants recapitulated this result ( Fig 1B ) . Food intake ( Fig 1C ) and locomotor rates [17] are indistinguishable between wild-type and gcy-33 mutants , suggesting that differences in the metabolism of fat stores underlies the fat phenotype of gcy-33 mutants . gcy-33 mutants are defective in the behavioral and neuronal responses to low oxygen , which are encoded by the BAG neurons [17 , 19] . In addition to the BAG neurons , GCY-33 is also reported to be expressed in the URX neurons [17 , 22] . To evaluate its site of action in regulating body fat , we restored gcy-33 cDNA in the gcy-33 mutants under its own promoter , and using heterologous promoters for the BAG and URX neurons . We observed robust partial rescue of body fat stores upon re-expression of GCY-33 under its endogenous promoter and in the BAG neurons , but did not observe meaningful rescue of body fat stores with its re-expression in the URX neurons ( Fig 1D and 1E ) . Incomplete rescue of gcy-33 mutants with exogenous cDNA has been previously observed for reasons that are currently not known [17] . Regardless , our experiments indicate that GCY-33 functions in the BAG neurons to regulate body fat stores . In accord with this hypothesis , we found that ablation of BAG neurons led to a profound decrease in body fat stores ( Fig 1F ) . We wished to evaluate whether gcy-33 mutants played a role in oxygen-dependent fat loss . Fasted wild-type adults initiate fat loss in an oxygen-dependent manner: wild-type animals exposed to 21% ( henceforth , 'high' ) oxygen metabolize approximately twice as much body fat as those exposed to 10% ( henceforth , 'low' ) oxygen ( Fig 2A and 2B ) . We previously showed that this effect is not regulated by changes in food intake or locomotor rates at high versus low oxygen [13] . Rather , it occurs because of a selective metabolic shift towards fat utilization that is differentially regulated by oxygen availability . Thus , animals exposed to high oxygen show greater fat loss than those exposed to low oxygen for the same duration . As expected , gcy-36 mutants show complete suppression of fat loss at high oxygen , consistent with the previously-observed role for the URX neurons in responses to high oxygen ( Fig 2B ) . Interestingly , we found that gcy-33 mutants showed the opposite phenotype to the gcy-36 mutants: they had greater fat loss at low oxygen than wild-type animals , and were indistinguishable from those exposed to high oxygen . This result is consistent with loss of BAG neuron activity in gcy-33 mutants [17] , and indicates a role for BAG neurons in suppressing fat loss at low oxygen ( Fig 2B ) . The observed effects of the BAG neurons on oxygen-dependent fat loss were specific to GCY-33 signaling , because null mutants of two other BAG-specific guanylate cyclases GCY-31 and GCY-9 [23] did not suppress oxygen-dependent fat loss ( Fig 2C ) . We were surprised to note that removal of the URX-specific sGC gcy-36 fully suppressed the phenotype of gcy-33 single mutants , such that the gcy-33;gcy-36 mutants resembled the gcy-36 single mutants alone ( Fig 2B ) . Because GCY-36 functions in the URX neurons and not in the BAG neurons [17] , this result suggested that the effect of GCY-33 signaling from the BAG neurons was dependent on GCY-36 signaling from the URX neurons ( Fig 2B ) . In keeping with the increased fat loss , gcy-33 mutants displayed increased energy expenditure that was fully suppressed by removal of gcy-36 ( Fig 2D ) . The increased fat loss of gcy-33 mutants under low oxygen ( Fig 2B ) was highly reminiscent of the gpa-8 mutants which we had previously studied ( Fig 2B; [13] ) . GPA-8 is a gustducin-like Gα protein which functions within the URX neurons themselves , and serves to inhibit GCY-36 . gpa-8 mutants have reduced fat stores in the fed state , show increased fat loss at low oxygen , and are indistinguishable from those fasted at high oxygen . gcy-33;gpa-8 double mutants resembled each single mutant alone , with no additive effects ( Fig 2B ) . We previously showed that the fat-regulatory property of gpa-8 mutants arises from increased constitutive activation of the URX neurons because of GCY-36 de-repression [13] , again resembling the gcy-33;gcy-36 mutants . Together , the data suggests that GCY-33 signaling from BAG neurons negatively regulates URX neurons and oxygen-dependent fat loss . To directly study the effects of GCY-33 signaling on URX function , we measured neuronal activity using Ca2+ imaging in living animals . We used the genetically-encoded calcium indicator GCaMP5K as a reporter for URX activity which has been optimized for greater sensitivity to threshold activation properties [13 , 24] . Wild-type animals bearing the GCaMP5K transgene expressed under the URX-specific promoter flp-8 showed robust calcium influx at 21% oxygen ( Fig 3A and 3C ) , as previously described [13 , 25] . We observed two properties of URX activation in gcy-33 mutants crossed into the Pflp-8::GCaMP5K transgenic line . First , at 21% oxygen there was an approximately 30% decrease in maximal activation of URX neurons in gcy-33 mutants compared to wild-type animals ( Fig 3B , 3D and 3E ) . Second , at 10% oxygen , we observed an approximately two-fold increase in baseline ( F0 ) fluorescence values in gcy-33 mutants compared to wild-type animals ( Fig 3F ) . Importantly , we verified that there was no observed difference in flp-8 promoter activity at 10% oxygen ( Fig 3G ) . When measuring URX peak responses to the oxygen upshift in absolute GCaMP5K fluorescence levels , no significant difference between wild-type animals and gcy-33 mutants was observed ( Fig 3H ) . The major effect of GCY-33 therefore lies in controlling Ca2+ concentrations in the URX neurons at 10% oxygen . These data were highly reminiscent of the effect of gpa-8 mutants on URX Ca2+ dynamics [13] . We next decided to evaluate URX signaling properties in the absence of BAG neurons . To this end , we crossed the Pflp-8::GCaMP5K transgenic line with the BAG ablation transgenic line , and measured URX properties in the double transgenic animals . Remarkably , at 21% oxygen there was a complete absence of URX responses in animals lacking BAG neurons ( Fig 4B , 4D and 4E ) . This effect occurred because of a 10-fold increase in median baseline ( F0 ) values at 10% oxygen ( Fig 4F ) . There were no changes in flp-8 promoter activity or other observable features of the URX neurons , including URX peak responses ( Fig 4G and 4H ) . Thus , BAG neurons , via gcy-33 signaling , inhibit resting URX properties at low ( 10% ) oxygen . BAG neurons are not known to form synaptic connections with the URX neurons . In addition , null mutants of the dense core vesicle-specific activator protein CAPS/UNC-31 that regulate neuropeptide secretion [26–28] blocked oxygen-dependent fat loss , whereas null mutants of the synaptic vesicle protein UNC-13 that regulates conventional neurotransmitter release [29 , 30] , did not ( S2 Fig ) . To explore a potential neuropeptide-based mechanism of communication from the BAG neurons to the URX neurons , we conducted a genetic screen of all available mutants of the neuropeptide gene families ( flp , nlp and ins gene families , 77/113 genes ) . Our top hit was a neuropeptide called FLP-17 , the canonical BAG-neuron-specific neuropeptide [31–33] ( Fig 5A ) . Interestingly , flp-17 null mutants were nearly identical to the gcy-33 mutants in oxygen-dependent fat loss . That is , relative to wild-type animals , fed flp-17 mutants had reduced fat stores , showed increased fasting-dependent fat loss at low oxygen , and were indistinguishable from those at high oxygen . In addition , gcy-33;flp-17 and flp-17;gpa-8 double mutants resembled either single mutant alone ( Fig 5A ) , suggesting that GCY-33 and FLP-17 from BAG neurons function in a linear pathway with GPA-8 signaling in URX neurons to inhibit fat loss at low oxygen . We generated a Pflp-17::flp-17GFP transgenic line that showed robust and selective expression in the BAG neurons , as expected ( Fig 5B ) . Interestingly , we observed punctate expression with GFP excluded from the nucleus , a pattern reminiscent of proteins destined for secretion [10] . To examine whether FLP-17 secretion is altered upon oxygen exposure , we measured FLP-17 secretion at low and high oxygen in these transgenic animals ( Fig 5C ) . Interestingly , we observed that FLP-17 secretion increased at low oxygen , and returned to baseline at high oxygen ( Fig 5C , 5D and 5E ) , in keeping with the activation of BAG neurons at low oxygen , and its inhibition at high oxygen [17] . Finally , we found that FLP-17 overexpression in BAG neurons significantly increased fat stores in the intestine ( Fig 5F ) . Together , these data show that in response to low oxygen , BAG neurons control the resting state of URX neurons via the FLP-17 neuropeptide . A GPCR called EGL-6 functions as the cognate receptor for the FLP-17 neuropeptide in the control of egg-laying [33] . Activation of the EGL-6 receptor inhibits egg-laying via its function in the HSN neurons , whereas loss-of-function mutants do not have an egg-laying phenotype . We found that null mutants of the egl-6 gene did not have overt defects in body fat stores , but were defective in oxygen-dependent fat loss indistinguishably from the flp-17 mutants: loss of egl-6 led to increased fat loss at low oxygen such that they resembled fat loss at high oxygen ( Fig 5G ) . flp-17;egl-6 double mutants resembled the egl-6 single mutants in all respects: in the fed state they did not show an appreciable difference in body fat stores , and showed increased fat loss in low oxygen , without additive effects . gcy-33;egl-6 and egl-6;gpa-8 double mutants also did not yield additive effects relative to each single mutant alone , suggesting that these genes function in a linear pathway to regulate oxygen-dependent fat loss ( Fig 5G ) . The decreased fat stores of flp-17 mutants was accompanied by increased respiration ( Fig 5H ) , whereas the egl-6 null mutants showed no differences , as predicted by their body fat phenotype in the fed state ( Fig 5G and 5H ) . Because our signaling pathway indicated communication from BAG neurons to URX neurons via the FLP-17 neuropeptide , we tested whether EGL-6 functions in the URX neurons . In egl-6 null mutants , we restored expression in the URX neurons and measured oxygen-dependent fat loss . Relative to non-transgenic animals , egl-6 re-expression in URX neurons fully restored oxygen-dependent fat loss ( Fig 6A ) , suggesting that EGL-6 functions in the URX neurons to limit fat loss at low oxygen . egl-6 expression had been noted in C . elegans head neurons [33] , and we observed clear and robust expression of egl-6 in the URX neurons ( Fig 6B ) . The egl-6 gain-of-function allele , n592 , displayed oxygen-dependent fat loss indistinguishably from wild-type animals , and completely suppressed the oxygen-dependent fat loss phenotype of flp-17 and gcy-33 mutants , but not of the gpa-8 mutants ( Fig 6C ) . Together , these data provide compelling evidence for a signaling pathway from the BAG neurons to the URX neurons for oxygen-dependent fat loss , via neuropeptide communication . For additional verification of the role of FLP-17 in controlling URX functions , we measured URX responses in the absence of flp-17 . Wild-type animals bearing the GCaMP5K transgene expressed under the URX-specific promoter flp-8 showed robust calcium influx at 21% oxygen ( Figs 7A and 3C ) , as previously described [13 , 25] . We observed two properties of URX activation in flp-17 mutants crossed into the Pflp-8::GCaMP5K transgenic line . First , at 21% oxygen there was an approximately 25% decrease in maximal activation of URX neurons in flp-17 mutants compared to wild-type animals ( Fig 7B , 7D and 7E ) . Second , at 10% oxygen , we observed an approximately 1 . 5-fold increase in baseline ( F0 ) fluorescence values in flp-17 mutants compared to wild-type animals ( Fig 7F ) . Importantly , we verified that there was no observed difference in flp-8 promoter activity at 10% oxygen ( Fig 7G ) . When measuring URX peak responses to the oxygen upshift in absolute GCaMP5K fluorescence levels , no significant difference between wild-type animals and flp-17 mutants was observed ( Fig 7H ) . The major effect of FLP-17 therefore lies in controlling Ca2+ concentrations in the URX neurons at 10% oxygen . These data were highly reminiscent of the effect of gcy-33 and gpa-8 mutants on URX Ca2+ dynamics [13] . Taken together , our results describe a neuropeptide-based signaling pathway from BAG to URX neurons that regulates the resting properties of URX neurons , and show that resting state modulation can have a major and lasting impact on metabolic homeostasis ( Fig 8 ) .
In this report , we show that the C . elegans oxygen-sensing neurons reciprocally regulate peripheral lipid metabolism . Under lower oxygen conditions , the BAG sensory neurons respond by repressing the URX neurons , which are tonic sensors of higher oxygen . Molecularly , repression occurs via a neuropeptide secreted from the BAG neurons , and its cognate receptor that functions in the URX neurons ( Fig 8 ) . Thus , BAG sensory neurons counterbalance the metabolic effect of the tonically active URX neurons . The physiological consequence of this repression is to limit fat utilization when oxygen levels are low . In previous work we had defined the role of the URX neurons in stimulating fat loss in response to high environmental oxygen ( Fig 8 , left panel ) . When oxygen levels are high and food supplies dwindle , URX neurons are activated via GCY-36 , ultimately leading to increased fat loss via the ATGL-1 lipase [13] . It has been noted that high oxygen coincides with dwindling bacteria , the main food source for C . elegans , because decreases in bacterial respiration would drop the local ambient concentration of oxygen below atmospheric ( 21% ) . In the present study we define a mechanism by which the URX neurons are held in the 'off state' under conditions of low oxygen , that is , via repression from BAG neurons . It has been appreciated for some years that URX neurons are tonic sensors of environmental oxygen [17] , suggesting that they must be turned off by an active repression mechanism , the molecular basis of which remained unknown . Our model suggests that BAG neurons in C . elegans play an essential role in mediating tonic repression of URX-neuron resting properties in the following way: the sensation of low oxygen concentrations by the guanylate cyclase GCY-33 in the BAG neurons initiates the release of the neuropeptide , FLP-17 , which is detected via the GPCR EGL-6 in the URX neurons . FLP-17/EGL-6 signaling leads to decreased basal activity of URX neurons via the GPA-8-mediated repression of GCY-36 . Thus , in low oxygen , URX neurons are retained in the 'off state' , ultimately limiting fat loss ( Fig 8 , right panel ) . Collectively , URX neurons are held in the off state in two ways . First , by internal sensing of body fat stores [13] and second , by the presence of food or low oxygen via the BAG neurons ( this study ) . We suggest that it is a combination of decreased GCY-33 activation in BAG neurons and increased GCY-36 activation in URX neurons that tunes the rate and extent of fat utilization in the intestinal cells . Work presented here uncovers the molecular basis for the conversion of oxygen sensation via the chemosensory system , to the regulation of lipid metabolism . Although the molecular nature of oxygen sensors in mammals is not yet known , oxygen concentration in blood , cerebrospinal fluid and other organs of the body vary widely [34–37] . We speculate that mammalian chemosensors of oxygen would similarly play a role in regulating lipid homeostasis in a manner that is independent of hypoxia-sensing [37] . The role of tonic repression is another interesting facet of sensory control that has emerged from our studies . Although tonic repression of neural activity is not well understood , our data indicate that neuropeptide-mediated repression of resting state is a robust mechanism to control the activity of sensory neurons . We suggest that tonic repression via neuropeptides and systemic hormones offers a broad and powerful mechanism to relay state information and control sensory and metabolic responses across the organs of the body . Our data indicate that the stimulation of fat loss is far more complex than what can be elicited simply by reduction in food intake or increased locomotion . Rather , environmental conditions , relayed by sensory systems , tune the rate and extent of fat loss which is in turn governed by complex genetic interactions between the nervous system and the intestine . We suggest that the mammalian counterparts to such regulatory pathways will be broadly informative for our understanding of fat metabolism .
C . elegans was cultured as described [38] . N2 Bristol , obtained from the Caenorhabditis Genetic Center ( CGC ) was used as the wild-type reference strain . The mutant and transgenic strains used are listed in S1 Table . Animals were synchronized for experiments by hypochlorite treatment , after which hatched L1 larvae were seeded on plates with the appropriate bacteria . All experiments were performed on day 1 adults . Promoters and genes were generated using standard PCR techniques from N2 Bristol worm lysates or cDNA and cloned using Gateway TechnologyTM ( Life Technologies ) . Promoter lengths were determined based on functional rescue and are outlined in S2 Table . All rescue plasmids were generated using polycistronic GFP . Transgenics were constructed by microinjection into the C . elegans germline followed by visual selection of transgenic animals under fluorescence . For the microinjections , 5–25 ng/μl of the desired plasmid was injected with 25 ng/μl of a Punc-122::GFP or Pmyo-3::mCherry coinjection marker and 50–70 ng/μl of an empty vector to maintain a total injection mix concentration of 100 ng/μl . In each case , 10–20 stable transgenic lines were generated . Two lines were selected for experimentation based on consistency of expression and transmission rate . For GCaMP5K transgenic animals , 5 ng/μl of Pflp-8::GCaMP5K was injected with 2 ng/μl of a Pflp-8::mCherry coinjection marker . Oil Red O staining was performed as described [13] . Within a single experiment , roughly 3 , 500 animals were fixed and stained , 100 animals were visually inspected on slides , following which 15–20 animals were imaged for each genotype/condition . All experiments were repeated at least 3 times . Black and white images of Oil Red O stained animals and fluorescent images were captured using a 10X objective on a Zeiss Axio Imager microscope . Lipid droplet staining in the first four pairs of intestinal cells was quantified by measuring background-subtracted pixel intensity after setting a standard threshold . Within each experiment , approximately 15–20 animals at the same stage of adulthood were quantified from each condition without selection bias . Images were quantified using ImageJ software ( NIH ) . For each group , 2 , 000 worms/10 cm plate were grown for 44 h at 25°C . After washing with PBS twice , worms were flash frozen in liquid nitrogen . Worms were homogenized in PBS containing 5% TritonX100 and proteinase inhibitor ( Thermo Scientific ) , and lipid was extracted using the TissueLyser II ( QIAGEN ) . Triglyceride content was measured using the EnzyChrom Triglyceride Assay Kit ( BioAssay Systems ) , and triglyceride levels were normalized to total protein , which was determined using the Pierce BCA Protein Assay ( Thermo Scientific ) . Food intake was measured by counting pharyngeal pumping , as previously described [39] . For each animal , the rhythmic contractions of the pharyngeal bulb were counted over a 10 s period under a Zeiss M2 Bio Discovery microscope . For each genotype , 10 animals were counted per condition and the experiment was repeated at least three times . The experiments were conducted as described [13] . For each strain , approximately 3 , 500 synchronized L1 larvae were seeded onto each of three plates . Worms were grown at 20°C for 48 h after which all plates were transferred to the bench top . Worms subjected to the fasting protocol were washed off the plates with PBS in 5 sequential washes over a 30-minute period to eliminate residual bacteria , and then seeded onto NGM plates without food . Worms were then subjected to a 2 . 5 h fasting period at either 21% or 10% oxygen . To establish the time course of fasting , pilot experiments were conducted at atmospheric ( 21% ) oxygen . The “21% fasted” plates were placed in a non-airtight container at room temperature . The “fed” control plates were placed in a similar but separate container . The “10% fasted” plates were placed in a custom-designed sealed acrylic oxygen chamber ( TSRI Instrumentation and Design Lab ) , fitted with inlet and outlet valves . The inlet valve was connected via bubble tubing to a pressurized oxygen and nitrogen pre-mixture containing 10% oxygen ( Praxair , Inc . ) , and the outlet valve was exposed to air . All plates were positioned right side up without lids . The sealed chamber was then perfused for 15 min with 10% oxygen . Following perfusion , both valves were closed . During the experiment , pressure inside the chamber was held constant , as judged by a gauge placed inside the oxygen chamber . The chamber was kept at room temperature for an additional 2 . 25 h , so that all fasted conditions remained off food for a total of 2 . 5 h following the washes . At the end of this period , worms from the respective conditions were collected for Oil Red O staining . Oxygen consumption rates ( OCR ) were recorded using the Seahorse XFe96 Analyzer ( Agilent ) . Worms were grown at 15°C for 24 h and then moved to 20°C for 48 h . Worms were washed with M9 buffer and transferred into a well in a 96-well plate at approximately 10 worms per well . The final volume per well was 200 μL of M9 . Worms were without food for 60 min prior to the first OCR recording . Respiration was measured according to the manufacturer's instructions . Oxygen consumption rates were normalized to worms per well . In all genotypes tested , we did not observe any changes in worm size , growth or developmental stage . We used a microfluidic chamber constructed with the oxygen-permeable poly ( dimethylsiloxane ) ( PDMS ) as described [17] . A Valvebank II ( AutoMate Scientific , Inc . ) was used to control input from two pressurized pre-mixtures of oxygen and nitrogen containing either 10% oxygen or 21% oxygen ( Praxair , Inc . ) . The gas flow rate was set to 0 . 26 psi at the outlet of the chamber as judged by a VWRTM traceable pressure meter . Immediately before imaging , individual day 1 adult animals were sequentially transferred to two unseeded plates . Individual C . elegans adults were then transported into the chamber in a drop of S Basal buffer containing 6mM levamisole ( Acros Organics B . V . B . A . ) via Tygon tubing ( Norton ) . Animals were constantly submerged in S Basal buffer while inside the chamber . After the animals were immobilized inside the chamber , GCaMP5K fluorescence was visualized at 40x magnification using a spinning disk confocal microscope ( Olympus ) using MetaMorphTM ( version 6 . 3r7 , Molecular Devices ) . Worms were pre-exposed to 10% oxygen for 5 min in the microfluidic chamber as described [17] . GCaMP5K fluorescence was recorded by stream acquisition for 2 min at a rate of 8 . 34 frames/second , with an exposure time of 20 ms using a 12-bit Hamamatsu ORCA-ER digital camera . Each animal was recorded once . GCaMP5K-expressing neurons were marked by a region of interest ( ROI ) . The position of the ROI was tracked using the “Track Objects” function in MetaMorphTM . An adjacent ROI was used to subtract background from the total integrated fluorescence intensity of the ROI . Data were analyzed using MATLAB ( MathWorks , Inc . ) . Fluorescence intensity is presented as the percent change in fluorescence relative to the baseline ( ΔF/F0 ) . F0 was measured in worms exposed to 10% oxygen during the first 9–13 seconds for each recording and calculated as an average over that period . All animals were day 1 adults at the time of imaging . The number of animals used for each condition is denoted in the figures . We used a microfluidic chamber constructed with the oxygen-permeable poly ( dimethylsiloxane ) ( PDMS ) as described [17] . A Valvebank II ( AutoMate Scientific , Inc . ) was used to control input from two pressurized pre-mixtures of oxygen and nitrogen containing either 10% oxygen or 21% oxygen ( Praxair , Inc . ) . The gas flow rate was set to 0 . 26 psi at the outlet of the chamber as judged by a VWRTM traceable pressure meter . Immediately before imaging , individual day 1 adult animals were sequentially transferred to two unseeded plates . Individual C . elegans adults were then transported into the chamber in a drop of S Basal buffer containing 6mM levamisole ( Acros Organics B . V . B . A . ) via Tygon tubing ( Norton ) . Pflp-17::flp-17GFP animals were constantly submerged in S Basal buffer while inside the chamber , with the oxygen concentration set to 21% . After the animals were immobilized inside the chamber , fluorescence was imaged at 40x magnification using a spinning disk confocal microscope ( Olympus ) using MetaMorphTM ( version 6 . 3r7 , Molecular Devices ) ( t = 0 ) . Immediately after the first image was taken , the oxygen concentration in the chamber was changed to 10% . With the worm exposed to 10% oxygen , fluorescence was imaged again after 10 min ( t = 10 ) and 30 min ( t = 30 ) . Immediately after the third image was taken , the oxygen concentration in the chamber was changed to 21% . After 10 min at 21% oxygen , fluorescence was imaged again ( t = 40 ) . Images were quantified using ImageJ software ( NIH ) . Wild-type animals were included as controls for every experiment . Error bars represent SEM . Student’s t-test , one-way ANOVA , and two-way ANOVA were used as indicated in the figure legends . | We now appreciate that the sensory nervous systems of complex multicellular animals play a profound role in influencing energy balance , and body fat stores . Understanding the precise molecular mechanisms and neuroendocrine pathways by which the nervous system controls metabolic tissues has remained a tremendous challenge . Using the genetically tractable nematode C . elegans , we can dissect the critical sensory modalities that influence body fat storage and its mobilization , and their mechanisms of action . In this study , we identify the molecular mechanisms by which a salient environmental feature , oxygen , regulates the magnitude of fat loss via the counterbalancing actions of the BAG and URX oxygen sensory neurons . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"biotechnology",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"caenorhabditis",
"oxygen",
"neuroscience",
"animals",
"animal",
"models",
"organisms",
"oxygen",
"metabolism",
"caenorhabditis",
"elegans",
"model",
"organisms",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"genotyping",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"lipids",
"genetically",
"modified",
"organisms",
"animal",
"cells",
"fats",
"chemistry",
"biological",
"tissue",
"molecular",
"biology",
"genetically",
"modified",
"animals",
"automated",
"fluorescent",
"genotyping",
"biochemistry",
"cellular",
"neuroscience",
"eukaryota",
"cell",
"biology",
"anatomy",
"adipose",
"tissue",
"neurons",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"metabolism",
"chemical",
"elements"
] | 2018 | Oxygen-sensing neurons reciprocally regulate peripheral lipid metabolism via neuropeptide signaling in Caenorhabditis elegans |
The timing of spiking activity across neurons is a fundamental aspect of the neural population code . Individual neurons in the retina , thalamus , and cortex can have very precise and repeatable responses but exhibit degraded temporal precision in response to suboptimal stimuli . To investigate the functional implications for neural populations in natural conditions , we recorded in vivo the simultaneous responses , to movies of natural scenes , of multiple thalamic neurons likely converging to a common neuronal target in primary visual cortex . We show that the response of individual neurons is less precise at lower contrast , but that spike timing precision across neurons is relatively insensitive to global changes in visual contrast . Overall , spike timing precision within and across cells is on the order of 10 ms . Since closely timed spikes are more efficient in inducing a spike in downstream cortical neurons , and since fine temporal precision is necessary to represent the more slowly varying natural environment , we argue that preserving relative spike timing at a ∼10-ms resolution is a crucial property of the neural code entering cortex .
The precision of neuronal spike trains is at the center of a fundamental debate in neuroscience as to what aspects of neuronal signaling are important in representing information in the brain . Individual neurons can have extremely precise and repeatable responses to the visual stimuli that strongly drive them ( down to 1-ms variability ) [1–5] , but they exhibit seemingly degraded temporal precision of firing activity in response to suboptimal stimuli [5–10] . In the presence of natural scenes , the activity of individual neurons is sparse [11] and precisely timed across repeated presentations of the visual stimulus , even though natural stimuli tend to vary on a time scale that is several times slower [12] . However , in most natural circumstances , the brain does not have access to multiple repetitions of the same identical stimulus , and , therefore , it is the precision of spiking across neuronal sub-populations on single trials that is ethologically relevant . While synchrony across neurons in the retina and visual cortex has been reported at various time scales , which can depend on the visual stimulus [10 , 13] , the temporal precision of the neural code directly entering primary visual cortex , and its dependence on the stimulus , are still unknown . We used natural visual stimuli to investigate the spike timing precision of populations of geniculate neurons that serve as the direct input to visual cortex . We show that the response of individual neurons is less precise across stimulus repetitions when luminance contrast is reduced . However , this reduction in the precision of spike timing is not observed at the level of the neuronal population . Therefore , spike timing precision in populations of geniculate neurons is relatively insensitive to global changes in visual contrast and remains constant on the order of ∼10 ms . Since closely timed spikes from either a single neuron [14] or several neurons [15] are more likely to induce a spike in the downstream cortical neuron to which they are projecting , and since fine temporal precision is necessary in representing the more slowly varying natural environment [12] , preserving the relative timing of spikes at a resolution of ∼10 ms may be a crucial aspect of the neural code entering primary visual cortex .
A short movie of a natural scene recorded from a “cat-cam” [16] was presented repeatedly to anesthetized cats while recording extracellular activity of multiple single units in the lateral geniculate nucleus ( LGN ) in vivo . To test how spike timing precision in individual cells and cell populations was affected by the properties of the visual stimulus , each group of cells was stimulated with both a high-contrast ( HC ) version and a low-contrast ( LC ) version of the movie . Figure 1A shows the firing activity of a typical ON-center geniculate cell in response to a 500-ms section of the movie presented at both high and low contrast . Each line in the raster plot corresponds to a single repeat and shows the spikes generated during one presentation of this movie section . The peri-stimulus time histogram ( PSTH ) shows the summed response across 64 repeated presentations . The geniculate neurons exhibited a typical pattern of response in which brief intervals of silence lasting 15 ms or more alternated with firing “events” [6] , or groups of closely spaced spikes lasting up to 100 ms ( median: 30 ms ) , which consistently occurred at approximately the same time on each movie presentation . A 63% reduction of the luminance contrast in the movie resulted in a 23% reduction in the neuronal firing rate , from 10 . 5 spikes/s at HC to 8 . 1 spikes/s at LC on average across cells ( see full distribution in Figure S1A ) , and in a latency increase of 3 . 4 ms ( Figure S1B ) . The firing rate reduction and latency increase are visible as a decrease in height and a shift in time in the PSTH events at LC in the particular example of Figure 1A . PSTH events are typically wider in duration at LC than at HC [17] ( as in the example rasters and PSTHs shown in Figure 1A , where the average event full width at half-height is 7 ms at HC , 11 ms at LC ) . Therefore , it has been assumed that suboptimal stimulation leads to a decrease in the temporal precision of the response . However , the widening of the PSTH events from the HC to LC condition could , in principle , arise from several sources , which have different functional implications: either an increase in the inter-spike intervals ( ISIs ) within the event on each trial ( Figure 1B , left ) , or an increased variability in the timing of the events themselves from trial to trial ( Figure 1B , right ) , or some combination of both factors . Further analyses were performed to test these two possibilities . First , over all cells , the distribution of ( within-trial ) ISIs was very similar at HC and LC . In Figure 1C , both distributions peak at 2 . 5 ms and have similar full width at half-height ( 8 ms at HC , 6 ms at LC ) . If ISIs were longer at LC than HC , the ISI distribution would be wider at LC than HC , which is the opposite of what we find . The fact that the ISI distribution at LC decays faster than at HC is simply due to the higher firing rate at HC , as is the case for Poisson processes ( e . g . , see [3] ) . Second , the timing of events showed significantly more across-trial variability ( or “jitter” ) at LC ( 10 . 9 ± 1 . 6 ms mean ± standard deviation ) than HC ( 8 . 2 ± 1 . 3 ms ) , with a 2 . 7-ms difference on average across cells ( paired t-test , p < 0 . 01 , n = 45 cells; see Figure 1D ) . These results suggest , therefore , that overall spike timing variability is primarily due to across-trial variability in the timing of the event as a whole . Ultimately , to compare within-cell and across-cell variability , correlation analysis is necessary . Therefore , we first re-examined and validated these single-cell precision findings in terms of correlation measures . First , we quantified the average PSTH event duration by measuring the width of the PSTH autocorrelation function , in which the temporal relationship between individual spikes is lost . We also quantified the temporal precision of individual spikes in a spike train by measuring the width of the spike autocorrelation function ( i . e . , the autocorrelation of the full spike train ) . As detailed in Figure S2 , the width of the PSTH autocorrelation includes two possible sources of spike timing variability ( within-trial and across-trial ) , whereas the temporal width of the spike autocorrelation function corresponds only to the within-trial variability . The PSTH autocorrelation and spike autocorrelation functions are shown in Figure 2A and 2B for four typical LGN cells at HC ( top ) and LC ( bottom ) , with the temporal widths in each condition indicated . The PSTH autocorrelation functions were significantly wider at LC than at HC on average ( Figure 2C; mean ± standard deviation: HC: 10 . 0 ± 2 . 5 ms; LC: 13 . 5 ± 3 . 1 ms; paired t-test , p = 5 × 10−17 , n = 45 ) , whereas spike autocorrelations did not show a significant difference in width between LC and HC ( Figure 2D; HC: 8 . 4 ± 2 . 8 ms; LC: 8 . 5 ± 3 . 8 ms; paired t-test , p = 0 . 46 , n = 45 ) , thus confirming that spike timing variability is primarily due to across-trial event-timing variability ( see Figure 1 ) . This result was found not only with natural movies , but also with spatiotemporal white noise visual stimulation ( Figure S3 ) , and was not related to the cell X or Y type ( Figure S4 ) nor to the occurrence of LGN bursts ( Figure S5 ) . These results indicate that decreasing the overall contrast increases the timing variability of groups of spikes ( events ) , but preserves the relative inter-spike timing precision within each group of spikes ( as illustrated in Figure 1B , bottom right ) , at a time scale on the order of ∼10 ms . The analyses reported above involved the global level of contrast in the full movie ( HC versus LC ) . To further elucidate the relationship between spike timing precision and contrast , we computed the local contrast experienced by each cell as the visual stimulus unfolds in time . Local values of spatiotemporal contrast ranged from 6% to 50% root-mean-squared ( RMS ) contrast ( see Methods ) . We classified each firing event as corresponding to one in four levels of local contrast: 6–13% , 13–20% , 20–34% , and 34–50% . For each of these contrast levels , we computed the PSTH autocorrelation and spike autocorrelation of each cell , as above . As shown in Figure 2E and 2F , the results presented above for two global levels of contrast ( HC and LC ) were confirmed with four levels of local contrast . The width of the PSTH autocorrelation significantly decreased as contrast increased ( Figure 2E; paired t-test; left to right pairwise comparisons: p = 1 × 10−6 , p = 1 × 10−7 , p = 2 × 10−4 , n = 45 cells ) , while spike autocorrelation did not show a significant difference in width across the four different levels of local contrast ( Figure 2F; paired t-test , p > 0 . 05 in all pairwise comparisons , n = 45 cells ) . If the relative timing of spikes is preserved at different levels of contrast in single cells , what does it imply across the population ? The activity of local groups of cells with neighboring receptive fields can be significantly correlated if the visual input itself has strong spatial and temporal correlations , as is the case with natural scenes [18–20] . Although it had been proposed that retinal and/or LGN neurons could remove these correlations through high-pass filtering achieved by lateral inhibition [21 , 22] , more recent neurophysiological studies suggest that the cells do not de-correlate their inputs [13 , 23] , and thus significant correlations from natural scenes remain present . The strength of pairwise correlation , defined as the area under the cross-correlation function ( in the HC condition ) , decreased with the distance between the receptive fields of two cells ( Figure 3A ) , following the decrease of spatial correlation strength in the visual stimulus ( Figure 3A , inset ) . We focused our analysis on pairs of cells that displayed sufficient cross-correlation in the HC condition ( n = 41 , see Figure 3A ) due to strong correlations in their visual input . Neurons with partially overlapped receptive fields ( Figure 3B ) typically receive similar visual input and therefore tend to share response events , as is evident in the two typical LGN X ON cells shown in Figure 3B and 3C . Both cells tended to fire during the same events , but there was some degree of timing variability ( see also Figure S6 ) . The increased variability in event timing in the LC condition for individual neurons , reported above ( see Figure 2 ) , could coexist with a range of effects across a population of neighboring cells , involving within-trial and across-trial variability in the relative timing of spikes from several cells . Since the above result indicates that within cells , event times are more variable across trials at lower stimulus contrast , we then tested the hypothesis that event timing across cells , both within trial and across trials , is also more variable at lower stimulus contrast . From the perspective of a downstream layer 4 V1 neuron receiving direct thalamic input , incoming spike trains that arrive simultaneously from a pair of LGN neurons may be represented by superimposing both spike trains into a single combined spike train . The ISI distribution for these combined spike trains is largely invariant to changes in contrast and peaks at 2 . 5 ms for both levels of contrast ( Figure 3D; full width at half-height: 6 ms at HC , 4 ms at LC ) , as was the case for single-cell spike trains ( Figure 1C ) . The across-cell event-time variability can be estimated by merging events from both cells that overlap in time and measuring the variability in the median time of the combined event ( Figure 3E; mean ± standard deviation of the event-time variability: 9 . 9 ± 1 . 0 ms at HC , 13 . 0 ± 1 . 4 ms at LC; n = 41 cell pairs ) . Across-cell event-time variability ranged between 8 and 19 ms , larger than but still on the same order of magnitude as that for single cells ( Figure 1D ) . Its average value was slightly higher in the LC than in the HC condition ( by 3 . 1 ms on average across cell pairs; p < 0 . 01 , n = 41 ) . However , it should be noted that this measure only indicates how the timing of combined firing events varies across repetitions of an identical stimulus . As argued above , it is the precision of spiking across neurons on single trials that is relevant for the neural population code . To quantify further the relative precision of spiking across the neuronal population , we computed cross-correlation in pairs of cells . While all pairs under study displayed stimulus-induced correlation , a few pairs also showed correlation on a finer time scale ( <1 ms ) , suggesting that they received common input from the same retinal ganglion cell [15] . Figure 4A and 4B shows the spike and PSTH cross-correlation functions for a pair of cells that shared input from the same retinal afferent ( as in 4/41 pairs ) and a pair of cells that did not . Importantly , since we are focusing on the neural representation of the visual scene rather than the details of the synaptic connectivity of LGN populations , our measure of spike correlation incorporates “signal” correlations ( inherited from correlations present in the visual stimulus ) as well as “noise” correlations ( arising from other sources such as shared input from a common retinal afferent ) . Both are integral components of the neural code in natural viewing conditions [24] and , taken together , reflect the relationships between the correlation structure of the visual scene and the functional properties of the local neuronal circuit . Across all pairs of cells under study , the temporal width of spike cross-correlation only showed a small difference between HC and LC ( Figure 4C; mean ± standard deviation: 14 . 7 ± 4 . 7 ms at HC , 15 . 7 ± 4 . 3 ms at LC; t-test , p = 0 . 05 , n = 41 pairs; see also control analyses in Figures S3–S5 ) . Moreover , the PSTH cross-correlation was very similar to spike cross-correlation ( Figure 4D ) . To investigate how spike timing precision of cell pairs is influenced by local visual contrast , we computed the spike and PSTH cross-correlation at different contrast levels: 6–13% , 13–20% , 20–34% , and 34–50% , as done previously for individual cells . The cross-correlation was based only on the events for which the local contrasts in both receptive fields were at the same level . Consistent with the results presented above , there was no trend in the width of spike and PSTH cross-correlations as a function of local contrast ( Figure 4E and 4F; some of the pairwise t-tests showed statistical significance , but not as a monotonic decrease in correlation width with increasing contrast ) . These results indicate that within-trial spike timing precision across cells is invariant to the change in contrast of the natural scene , despite the increased variability in event timings for individual cells across trials with decreasing contrast . As evident in Figures 2F and 4E , the spike cross-correlation obtained from pairs of neurons was consistently wider than the spike autocorrelation obtained from each individual neuron . The difference in width was 8 ms on average ( two-sided Wilcoxon rank sum test , p < 1 × 10−6 for each of four contrast levels , n = 45 cells , n = 41 cell pairs ) and was also found on a pair-by-pair basis . In almost all cell pairs , cross-correlation width was significantly greater than the width of the autocorrelation functions of both cells , as shown in Figure 5A ( see Figure S8 for the case of pairs lying close to the unity line , i . e . , with similar within-cell and across-cell precision ) . This finding indicates that spike timing precision was coarser in neuronal pairs than in individual neurons . This decrease in precision can be explained by the fact that , in general , the events are not perfectly aligned across both cells , as illustrated in Figure 5B . Even if cells have wider PSTH events at LC than HC , the overall increase in event time variability from HC to LC by 3 ms is small in the face of pairwise variability , which is on average 8 ms greater than single-cell variability . Another way to compare pairwise variability with contrast-based variability is by computing , for each shared event , the difference in event time between both cells ( and its variability ) and the difference in event time between the HC and LC condition ( and its variability ) . As shown in Figure 5C , the difference in event times between two cells is more variable ( i . e . , has a wider distribution ) than the difference in event times between HC and LC . The standard deviations of the distributions across all events are 16 ms ( HC ) and 18 ms ( LC ) across cells , and only 11 ms across contrast ( n = 4 , 205 events ) . In other words , the variability in event timing across cells is approximately 1 . 5 times larger than the variability in event timing across levels of contrast . Therefore , in the face of across-cell variability , the smaller changes in variability due to changes in contrast are negligible . Thus , spike timing precision across most neighboring cells is relatively insensitive to contrast .
In response to movies of natural scenes , spike timing precision across LGN relay cells remained on the order of ∼10 ms , irrespective of contrast . The absolute timing of LGN firing events changed from trial to trial , and more so at low contrast than at high contrast , but the relative timing of spikes occurring in the same trial was insensitive to changes in stimulus contrast—not only within cells but also across correlated neighboring cells . While it is well known that the response properties of single cells are strongly modulated by contrast adaptation , which has effects including slower temporal dynamics and increased gain and selectivity at lower contrast [17 , 25 , 26] , our results indicate that the temporal precision of the LGN population code is globally maintained in the face of a reduction in contrast . Interestingly , while in individual neurons PSTH autocorrelations were consistently wider at low contrast than high contrast ( Figure 2C and 2E ) , the width of the PSTH ( and spike ) cross-correlation was independent of contrast ( Figure 4C , 4E , and 4F ) . This surprising result can be explained by the large variability in event timing across populations of neurons , which is about 1 . 5 times larger than the variability in event timing caused by changes in contrast ( Figure 5C ) . A related finding is that while in individual neurons , the PSTH autocorrelations were consistently wider than the spike autocorrelations , in cell pairs , PSTH and spike cross-correlation had very similar widths ( Figure 4D and 4F ) . This finding suggests that correlations between cells arose mostly from correlations present in the visual input in our experimental conditions , and that neural “noise” ( or across-trial variability arising from intrinsic properties of the system ) shows little correlation across neurons ( Figure S7 ) . Nevertheless , it should be noted that the presence of weak , pairwise noise correlations does not rule out the possibility of stronger , higher-order correlations at the population level [27–29] . Downstream from the LGN , the influence of stimulus contrast on the timing of spikes across V1 cells has only been recently addressed . Spike timing precision across cells in anesthetized macaque V1 reportedly decreased for low-contrast grating stimuli [10] , whereas we found that it was contrast-independent in the cat LGN for natural stimuli . Further , a recent study in cortical slices suggested that the degree of noise correlation between two neighboring cortical cells increased with firing rate [30] , unlike the thalamic signal and noise correlations reported here that were invariant to contrast-driven changes in firing rates . These discrepancies may be attributable to differences in the experimental preparations and in the visual stimuli , or they could be explained by specific contrast adaptation mechanisms that occur in cortex but not in precortical areas . For example , in the peripheral auditory system , adaptation to a constant stimulus reduces the firing rate but does not impair spike timing precision [31] , in a similar fashion as what we found in the visual thalamus . Preserving spike timing across cells at a ∼10-ms resolution may be a crucial aspect of the neural population code in natural conditions , given that the representation of spatiotemporally varying natural scenes requires a finer temporal precision than the time scale of the visual stimulus [12] . Furthermore , a 10-ms temporal resolution could facilitate “temporal coding” under the hypothesis that the neural representation of sensory information relies on specific temporal patterns of spikes [32–34] . However , the existence or preservation of specific temporal patterns is beyond the scope of the present study . Downstream from the thalamus , spike timing precision may well vary along the visual pathway . Single-cell studies found that trial-to-trial variability is similar in the LGN and V1 when a V1 cell is presented with its preferred stimulus , but that V1 cells become more variable for suboptimal stimuli [5 , 8 , 35] . In the presence of natural movies , which combine optimal and nonoptimal stimulation for each cell , recent studies in primate V1 indicate that some visual information is present in the phase of local field potentials at low frequency ( <12 Hz ) [36] and that power associated with spiking activity is only informative at frequencies under 12 Hz [37] . Therefore , the relevant time scale in V1 is probably on the order of tens of milliseconds , only slighter longer than what we found in the LGN . The small increase in variability in V1 trial-to-trial spike timing compared to the LGN [35] may be explained by nonlinearities in the spiking mechanism and may coexist with lower variability in V1 membrane potential [38] . It is also possible that temporal precision is higher between cortical cells receiving input from geniculate cells that share a common retinal afferent , in a divergent–convergent pattern of connectivity . In any case , it is difficult to relate these previous results to population coding . The degree of spike timing precision across V1 cells , especially in natural viewing conditions , is not well quantified , and how it would be affected by contrast or other variables is unknown . Further studies are needed to elucidate how the functional architecture of the thalamocortical circuit constrains spike timing precision across cells and how it affects the neural code entering V1 . Preserving synchrony across cells could have a number of functional advantages . Synchronous spikes from several thalamic neurons are reportedly needed to drive cortical cells to threshold [15 , 39] . Recent studies have suggested that the cortical response is sensitive to the timing of thalamic inputs and that the “window of opportunity” for integration of excitatory inputs at the thalamocortical synapse remains unchanged in the face of adaptation [40] . Therefore , the relative timing of spikes in thalamic neurons could be an important aspect of the population neural code entering primary sensory cortices and could benefit from being insensitive to some properties of the sensory world while maintaining sensitivity to other , presumably more interesting , features .
Single-cell activity was recorded extracellularly in the LGN of anesthetized and paralyzed cats using a seven-electrode system . Four animals were used for a total of ten electrode penetrations . Surgical and experimental procedures were performed in accordance with United States Department of Agriculture guidelines and were approved by the Institutional Animal Care and Use Committee at the State University of New York , State College of Optometry . As described in [41] , cats were initially anaesthetized with ketamine ( 10 mg kg−1 intramuscular ) followed by thiopental sodium ( 20 mg kg−1 intravenous during surgery and at a continuous rate of 1–2 mg kg−1 h−1 intravenous during recording; supplemented as needed ) . A craniotomy and duratomy were performed to introduce recording electrodes into the LGN ( anterior , 5 . 5; lateral , 10 . 5 ) . Animals were paralyzed with atracurium besylate ( 0 . 6–1 mg kg−1 h−1 intravenous ) to minimize eye movements , and were artificially ventilated . Geniculate cells were recorded extracellularly from layer A of LGN with a multielectrode matrix of seven electrodes [42] . The multielectrode array was introduced in the brain with an angle that was precisely adjusted ( 25–30 degrees antero-posterior , 2–5 degrees lateral-central ) to record from iso-retinotopic lines across the depth of the LGN . A glass guide tube with an inner diameter of ∼300 μm at the tip was attached to the shaft probe of the multi-electrode to reduce the inter-electrode distances to approximately 80–300 μm . Layer A of LGN was physiologically identified by performing several electrode penetrations to map the retinotopic organization of the LGN and center the multielectrode array at the retinotopic location selected for this study ( 5–10 degrees eccentricity ) . Recorded voltage signals were conventionally amplified , filtered , and passed to a computer running the RASPUTIN software package ( Plexon ) . For each cell , spike waveforms were identified initially during the experiment and were verified carefully off-line by spike-sorting analysis . Cells were classified as X or Y according to their responses to counterphase sinusoidal gratings . Cells were eliminated from this study if they did not have at least 2 Hz mean firing rates in response to all stimulus conditions , or if the maximum amplitude of their spike-triggered average in response to spatiotemporal white noise stimuli was not at least five times greater than the amplitude outside of the receptive field area . For each cell in the main experiments , visual stimulation consisted of 128–240 repeats of one of two short movies of natural scenes taken from “cat-cam” movies recorded from a small camera mounted on top of a cat's head while roaming in grasslands and forests [16] . As in [17] , to improve temporal resolution , movies were interpolated by a factor of two ( from 25 to 50 Hz ) using commercial software ( MotionPerfect , Dynapel Systems Inc . ) and then presented at 60 frames per second , i . e . , at 1 . 2× speed . Following interpolation , the intensities of each movie frame were rescaled to have a mean value of 125 ( where the full range of intensity values was 0–255 ) . Each movie spanned 48 × 48 pixels at an angular resolution of 0 . 2 degree per pixel . The first movie ( presented to 28 of the cells included in the final analysis ) was 750 frames and lasted 12 . 5 s , while the second movie ( presented to the remaining 17 cells ) was 600 frames long and lasted 10 s . The stimuli were presented at 60 frames per second with a 120-Hz monitor refresh rate , such that each frame was displayed twice . Each movie was repeated 64–120 times at each of two global levels of luminance contrast: 0 . 4 ( high contrast , or HC ) and 0 . 15 ( low contrast , or LC ) RMS contrast [17] . In addition to “cat-cam” natural movies , as a control for each cell we also used visual stimulation consisting of spatiotemporal binary white noise shown at high contrast ( 0 . 55 RMS contrast ) and low contrast ( 0 . 20 RMS contrast ) . The spatial resolution and refresh rate of the white noise stimulus were the same as those of the natural scene movies . Each cell in the reported data was stimulated with the natural scenes movies as well as the white noise stimuli with an equal number of repeats ( 120 repeats at each level of contrast for 28/45 cells , 64 repeats at each level of contrast for 17/45 cells ) . For each cell , the spatiotemporal receptive field was estimated by standard spike-triggered-average techniques based on spatiotemporal white noise stimuli [43 , 44] . The spatial receptive field was fitted with a difference of two-dimensional Gaussians . The distance between receptive fields was defined as the distance between the centers of the Gaussians . The diameter of each receptive field was estimated as the average length of the major and minor axes of the one–standard deviation ellipse that defines the receptive field center . The overlap between two receptive fields was evaluated as the normalized dot product of the two receptive fields , computed after each receptive field had been normalized so that its dot product with itself was one [45 , 46] . For each cell at each level of contrast ( HC or LC ) , a single PSTH was computed as the cumulative response of the cell over all 64–120 repeats of the same short movie . Each PSTH was therefore 10 or 12 . 5 s long , depending on the duration of the stimulus presented to the cell . ISIs were computed as the time intervals between consecutive spikes; in the case of pairs of cells , we merged the spike trains from both cells and computed the ISIs from the combined spike train . Bursts were identified as groups of spikes separated from each other by 4 ms or less , where the first spike is preceded by a period of silence of 100 ms or more [47–49] . The degree of burstiness exhibited by each neuron was defined as the percentage of spikes belonging to a burst . Previous studies typically define temporal precision of single neurons as the standard deviation of the spike times within an identified event across trials [6–9 , 31 , 35] . In this study , we first defined a related measure which is the ( temporal ) width of the central peak in the PSTH autocorrelation [50] . The width of PSTH events and the width of the PSTH autocorrelation function are directly related , by a factor of √2 in the Gaussian approximation . In computing the PSTH ( and its autocorrelation ) , all spike trains that the cell produced in response to multiple repeats of an identical stimulus were collapsed into one “lumped” spike train ( i . e . , a PSTH with a 1-ms bin size , of the same duration as a single presentation of the movie , i . e . , 10 or 12 . 5 s ) . In the PSTH autocorrelation measure , the relative timing of spikes within a given trial or across all trials were confounded . To investigate within-trial temporal precision , we therefore computed a different measure: the width of the central broad peak in the spike autocorrelation , which we defined as the autocorrelation function of the full ( several minutes long ) spike train without collapsing the trials together [51 , 52] . Although analysis of single cells was a necessary first step , the primary focus of this study was on spike timing variability across cells . Two definitions of cross-correlation were used: spike cross-correlation [52 , 53] and PSTH cross-correlation , which is the cross-correlation between two PSTHs . Spike cross-correlation width gives the spike timing variability across cells within each trial . PSTH cross-correlation has a different meaning: it is approximately equivalent to the “shuffled” or “shifted” spike correlation , in which each spike train of one cell is paired with a spike train of the other cell recorded during a different repeat of the same stimulus . The PSTH cross-correlation averages correlations from all possible pairwise combinations of repeats ( actually including the non-shuffled one , which is only one in thousands of combinations and therefore has a negligible contribution ) . All four types of correlation functions ( spike or PSTH , auto- or cross-correlation ) were made analogous to Pearson's correlation coefficient by ( i ) subtracting the product of the average firing rates , and ( ii ) dividing by a normalization factor ( see below ) , such that correlation could take values between −1 and +1 . To determine the existence of a central peak or trough in a correlation function , we found the Gaussian function that best fit the central ±100 ms , in a least-mean-square sense . The standard deviation of this Gaussian provides a measure of the correlation width . In the case of autocorrelation , the height Ai of the best-fitting Gaussian was measured for each cell i and was subsequently set to 1 to normalize the autocorrelation function . In the case of cross-correlation between cells i and j , the best-fitting Gaussian was normalized by a factor of , where Ai and Aj are the heights of each respective autocorrelation function before normalization . The area under the Gaussian curve after normalization was used to define the strength of the cross-correlation between two neurons . Inclusion criterion for pairs: A pair of cells was included in the final pairwise analysis if its spike cross-correlation function peaked at a value of 0 . 065 or higher , an arbitrary threshold below which the cross-correlation function could not be well fitted by a Gaussian function . For all pairs of pixels corresponding to the receptive field centers of pairs of cells , we measured the correlation function between both time series ( i . e . , the time series of the intensity values of each pixel across all frames of the movie ) . Correlation strength was defined as the area under the Gaussian curve that best fit the cross-correlation function . The resulting spatial profile of correlation in the visual input , i . e . , the graph of correlation strength as a function of the distance between two pixels , was fitted ( in the least-mean-square sense ) to an exponential function with a negative exponent , which is the form expected for spatial correlations in a signal with a power spectrum decreasing as 1/f2 with spatial frequency . Single-cell event analysis: PSTH “events” were first defined in the PSTH at HC as times of firing interspersed with periods of silence lasting at least 20 ms . If the standard deviation of all spike times constituting an event was less than 20 ms , an attempt was made to break up the event into several events , a procedure in which the spikes were fitted to a mixture-of-Gaussians model using the Expectancy Maximization ( EM ) algorithm for maximum likelihood [54] . PSTH events at LC were then defined by aligning LC spikes to existing HC events if possible , with a preference for an HC event that occurred earlier rather than after the LC spike ( since it is known that spikes tend to be more delayed at LC than HC ) . If no corresponding HC event was found , a new event was created at LC , with a corresponding empty event at HC . The timing of an event on a given repeat was defined as the median time of all spikes composing this event . For each event at a given contrast level , the event time variability was the standard deviation of the timing of the event across repeats . We computed for each cell its average event time variability across all events . Pairwise event analysis: Starting from the single-cell event analysis above , each event from the first cell was matched to one or several events in the second cell with which it overlapped in time . If several events in one cell could be matched to a single event in the other cell , these events were merged into one . The list of all events that could be matched across the two cells constituted the list of “shared events . ” For each shared event at a given contrast level , the event time variability was the standard deviation of the timing of the event across repeats and across both cells . We computed for each cell pair its average event time variability across all events . Event-by-event analysis of event time difference , within cells and across cells: For all pairs ( cell A and cell B ) , for each pairwise event that existed in the four cases ( cell A at HC , cell A at LC , cell B at HC , and cell B at LC ) , we computed within-cell HC-LC event time difference as the average event time at LC minus the average event time at HC , for each of the two cells ( cell A and cell B ) . In other words , we hold the cell fixed and varied across two contrast levels . We also computed across-cell event time difference at a given contrast level ( HC or LC ) as the average event time for cell A minus the average event time for cell B . In this case , we hold contrast fixed and varied across two cells . Therefore , each pairwise event yielded four different data points ( 2 × 2 ) to compare the distributions of across-cell and within-cell event time difference , as shown in Figure 5C . For each cell , we computed the local value of spatiotemporal contrast as follows . For each firing event determined as above , we identified the smallest rectangle in the image that encompassed the cell's receptive field ( e . g . , 3 × 4 pixels ) and extracted from the movie the luminance values of these pixels at the six previous frames . Six movie frames at 60 Hz correspond to a duration of ∼100 ms , matching the temporal kernel of the cells . The RMS contrast of this spatiotemporal patch of the movie was computed as the standard deviation over all the corresponding pixel values ( e . g . , 3 × 4 × 6 values ) . In the LC movie , local contrast values in the 45 cells ranged from 6–20% RMS contrast . In the HC movie , they ranged from 14–50% . For each cell , each firing event ( in either the HC or LC condition ) was assigned one in four levels of local contrast: 6–13% , 13–20% , 20–34% , or 34–50% . Correlation analysis was then performed as described above on small sections of data corresponding to individual events . We restricted the cross-correlation analysis to the firing events for which both cells experienced a value of local contrast that fell into the same range ( out of the four ranges defined above ) . | Neurons convey information about the world in the form of trains of action potentials ( spikes ) . These trains are highly repeatable when the same stimulus is presented multiple times , and this temporal precision across repetitions can be as fine as a few milliseconds . It is usually assumed that this time scale also corresponds to the timing precision of several neighboring neurons firing in concert . However , the relative timing of spikes emitted by different neurons in a local population is not necessarily as fine as the temporal precision across repetitions within a single neuron . In the visual system of the brain , the level of contrast in the image entering the retina can affect single-neuron temporal precision , but the effects of contrast on the neural population code are unknown . Here we show that the temporal scale of the population code entering visual cortex is on the order of 10 ms and is largely insensitive to changes in visual contrast . Since closely timed spikes are more efficient in inducing a spike in downstream cortical neurons , and since fine temporal precision is necessary in representing the more slowly varying natural environment , preserving relative spike timing at a ∼10-ms resolution may be a crucial property of the neural code entering cortex . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience"
] | 2008 | Timing Precision in Population Coding of Natural Scenes in the Early Visual System |
In C . elegans , efficient RNA silencing requires small RNA amplification mediated by RNA-dependent RNA polymerases ( RdRPs ) . RRF-1 , an RdRP , and other Mutator complex proteins localize to Mutator foci , which are perinuclear germline foci that associate with nuclear pores and P granules to facilitate small RNA amplification . The Mutator complex protein MUT-16 is critical for Mutator foci assembly . By analyzing small deletions of MUT-16 , we identify specific regions of the protein that recruit other Mutator complex components and demonstrate that it acts as a scaffolding protein . We further determine that the C-terminal region of MUT-16 , a portion of which contains predicted intrinsic disorder , is necessary and sufficient to promote Mutator foci formation . Finally , we establish that MUT-16 foci have many properties consistent with a phase-separated condensate and propose that Mutator foci form through liquid-liquid phase separation of MUT-16 . P granules , which contain additional RNA silencing proteins , have previously been shown to have liquid-like properties . Thus , RNA silencing in C . elegans germ cells may rely on multiple phase-separated compartments through which sorting , processing , and silencing of mRNAs occurs .
RNA silencing is an anciently conserved pathway that regulates gene expression in most eukaryotes . Key to this pathway are members of the Argonaute protein family , which bind a diverse set of small regulatory RNAs , ranging from ~18–30 nucleotides in length . This small RNA-Argonaute complex regulates fully or partially complementary mRNAs at the level of transcription , translation , and stability [1 , 2] . By regulating both endogenous and foreign RNAs , small RNAs maintain proper gene expression , silence deleterious RNA products , and play critical roles in development , chromosome segregation , transposon silencing , fertility , and viral defense [2 , 3] . Small RNAs can be generated through a variety of mechanisms . Long double-stranded RNAs ( dsRNAs ) , which can be produced from exogenous or endogenous sources , are cleaved into small-interfering RNAs ( siRNAs ) by the RNase III-like enzyme Dicer [4 , 5] . When these siRNAs are produced from a primary dsRNA source such as viruses or convergent transcription , they are referred to as primary siRNAs . In many organisms , including C . elegans , plants , and fungi , RNA-dependent RNA polymerases ( RdRPs ) use mRNAs targeted by primary siRNAs as a template to amplify and maintain the silencing signal [6–9] . RdRPs can function either by synthesis of additional long dsRNA and coordinated cleavage by Dicer into secondary siRNAs , or , as in C . elegans , RdRPs can directly synthesize secondary siRNAs antisense to the target mRNA [10–13] . In previous work , we characterized MUT-16 as a protein essential for RNA silencing and nucleation of a nuclear pore-associated RNA silencing compartment in C . elegans germ cells called Mutator foci [14] . MUT-16 recruits many other proteins required for RNA interference and endogenous small RNA biogenesis to Mutator foci , including the nucleotidyl transferase MUT-2 , the 3’-5’ exonuclease MUT-7 , the DEAD-box RNA helicases MUT-14 and SMUT-1 , and two proteins of unknown function , RDE-2 and MUT-15 [14 , 15] . Additionally , the RNA-dependent RNA polymerase RRF-1 and the Zc3h12a-like ribonuclease RDE-8 localize to Mutator foci , however their dependencies for localization were not known [14 , 16] . No proteins have yet been identified that are required for MUT-16 localization , suggesting that MUT-16 may be the primary mediator of Mutator foci formation . Mutator foci are considered hubs of siRNA amplification because mutations in mut-16 or any of the associated Mutator complex proteins ( mut-2 , mut-7 , mut-14 smut-1 , mut-15 , rde-2 , rde-8 , or rrf-1 ) result in a substantial loss of the RdRP-dependent secondary siRNAs [7 , 9 , 14–17] . Furthermore , Mutator foci reside adjacent to another ribonucleoprotein ( RNP ) granule , the P granule , which contains additional proteins associated with RNA silencing and mRNA decay [14] . Notably , the Argonaute protein WAGO-1 , which binds secondary siRNAs synthesized in the Mutator complex , localizes to and presumably silences mRNAs complementary to its bound siRNAs in the P granule [7 , 18] . It is unclear how mRNAs targeted by primary Argonaute proteins get trafficked to the Mutator foci or how siRNAs generated in the Mutator foci end up bound by WAGO-1 in the P granule . However , the close juxtaposition between the P granule , the Mutator foci , and the nuclear pore suggests a role for sorting RNAs and proteins between compartments during the multi-step process of mRNA recognition and RNA silencing . Little is known about how MUT-16 nucleates Mutator foci , but one clue comes from its protein sequence . MUT-16 is highly enriched in disorder-promoting amino acids such as glutamine and proline [14 , 19] . Intrinsically disordered regions ( IDRs ) , which lack predicted structure , are often found in proteins linked to the formation of RNA granules ( i . e . the nucleolus , P bodies , stress granules , and germ granules ) [20–27] . Transient interactions between IDR-containing proteins can be a driving force for RNA granule formation through protein condensation and liquid-liquid phase separation , which occurs when proteins and RNAs self-organize to form a distinct compartment with liquid-like characteristics , but separate from the cytoplasm or nucleoplasm . This liquid droplet or granule , sometimes referred to as membrane-less organelle , can readily exchange components with the surrounding cytoplasm or nucleoplasm [21 , 24 , 26 , 28] . Thus , liquid-liquid phase separation can facilitate reactions by increasing the local concentration of proteins or RNA , as has been proposed for the nucleolus [26] . Alternatively , condensation can segregate certain factors away from the cytoplasm , for example sequestration of mRNAs and translational repressors into P bodies [29] . Here we demonstrate that the assembly of the Mutator complex and the formation of Mutator foci are mediated by the intrinsically disordered protein MUT-16 . MUT-16 acts as a scaffold to recruit RRF-1 and other Mutator complex proteins , and it promotes foci formation through its C-terminal region . We further show that Mutator foci depend on weak hydrophobic interactions for their formation and form in a concentration and temperature-dependent manner . Mutator foci also recover rapidly after photobleaching . Thus , our data suggest that Mutator foci are phase-separated compartments with liquid-like properties that associate with P granules and nuclear pores in the cytoplasm of germ cells to promote RNA silencing .
In previous work , we observed that MUT-16 was both Q/N-rich and P-rich , predominantly in its C-terminal region and contains no other conserved domains [14] . Because amino acid sequences containing low complexity regions , such as Q/N-rich or P-rich regions , are often associated with disorder , we sought to identify regions of intrinsic disorder within the MUT-16 protein sequence using IUPred [30 , 31] . MUT-16 has a short IDR near the N-terminus ( first ~100 amino acids ) and a much longer IDR comprising approximately 60% of the protein ( Fig 1A ) . In total , more than 70% of the protein is predicted to be unstructured . To determine if the intrinsically disordered nature of MUT-16 is a conserved feature of this protein , we used IUPred to predict IDR within the orthologs of MUT-16 , in C . remanei , C . briggsae , and C . japonica . Despite the relatively low sequence conservation between orthologs , particularly in the C-terminal half of the protein [14] , the conservation of IDRs was striking ( Fig 1A ) . Like in C . elegans , the C . remanei , C . briggsae , and C . japonica MUT-16 proteins have a short IDR near the N-terminus and much longer IDR comprising the majority of the middle to C-terminal portions of the proteins . To determine if the high incidence of Q/N/P residues is distributed throughout the IDR of MUT-16 or if it is restricted to a subset of the IDR , we counted the number of each of these residues in a sliding window of 100 amino acids , starting at position one and shifting 10 residues at a time . We observed a prominent peak of glutamine which overlapped with and was somewhat preceded by a prominent peak of proline ( Fig 1B ) . These peaks spanned only a subset of the C-terminal IDR . Like the conservation of IDR , the pattern of enrichment of Q/N/P residues across MUT-16 orthologs in C . remanei , C . briggsae , and C . japonica is strikingly similar and suggestive of a functional role for Q/N/P residues ( S1 Fig ) . Because IDRs have been shown to promote phase separation of proteins into liquid-like droplets [20–27] , and MUT-16 forms foci in vivo , we sought to establish whether some or all of the MUT-16 IDR was required for Mutator foci formation . We also sought to identify regions of MUT-16 that directly recruit other Mutator complex proteins ( Fig 1C ) . To this end , we generated a series of small deletions of the MUT-16 protein using CRISPR genome editing ( Fig 1D and 1E ) . Each deletion removes between 61 and 141 residues ( 6–13% ) of MUT-16 . For simplicity , we refer to the deletions as ΔA through ΔL . We initially planned to make twelve deletions but due to technical constraints , the ΔH and ΔI deletions were combined to make the single , slightly larger ΔH-I deletion . We additionally generated two large deletions that remove most or all of the large IDR of MUT-16 , which we will refer to as ΔE-I or ΔE-K . A strain containing mut-16 ( pk710 ) , a null mutation in the mut-16 gene , is defective in germline and somatic exogenous RNAi [14 , 17 , 32] . To determine the severity of RNAi defects caused by each mut-16 deletion , we tested the response of each mutant strain to dsRNA targeting the germline gene pos-1 , which causes embryonic lethality . We also tested their response to the somatic genes nhr-23 , which causes larval arrest , and lin-29 , which causes the intestine and gonad of the animal to rupture through the vulva at the larval to adult transition . Surprisingly , our RNAi assays revealed that only mut-16 deletions ΔC , ΔF , ΔH-I , and ΔL have a significant impact on somatic RNAi ( Fig 2A and 2B ) . The remaining deletions had only mild or no RNAi defect . In contrast , all deletions in mut-16 had defects in germline RNAi , though deletions in ΔA and ΔG had more modest effects ( Fig 2C ) . These data reveal that the majority of the MUT-16 protein is necessary for robust germline RNAi . In contrast , some regions of MUT-16 are dispensable for the response to at least some somatic RNAi clones , suggesting that the soma may be more resilient to mild perturbations in MUT-16 . Each of the mut-16 deletions were generated in the mut-16::mcherry::2xHA background . To determine which of these regions is required to promote foci formation , we examined the localization of mut-16 in each deletion background by live imaging ( Fig 3A–3E ) . Most deletions , including ΔA , ΔB , ΔD , ΔE , ΔF , ΔG , and ΔH-I , formed foci with similar intensity to the control ( full-length ) strain ( Fig 3A and 3B ) . The ΔC mutation still formed Mutator foci , but had an overall reduced fluorescence ( Fig 3A and 3C ) . It is unclear whether the reduced fluorescence intensity observed in the ΔC mutation is due to reduced expression or stability of MUT-16 specifically in the germline , but the overall MUT-16 protein levels are similar to full-length MUT-16 ( Fig 3F ) . In contrast , ΔJ , ΔK , and ΔL displayed severely disrupted foci , though the overall expression of cytoplasmic MUT-16 was not reduced ( Fig 3A and 3D ) . We also examined the two large deletions , ΔE-I and ΔE-K , which remove most of the IDR . Surprisingly , ΔE-I , which removes 387 amino acids ( ~37% of the protein ) and encompasses a substantial portion of the IDR , does not affect MUT-16 localization , indicating that a large portion of the IDR is dispensable for foci formation . In contrast , ΔE-K does disrupt MUT-16 localization , which is not unexpected given that it includes J and K regions , which individually disrupted foci formation . All deletions were expressed at similar or higher levels than the full-length strain ( Fig 3F ) , indicting that the reduced level of foci in ΔJ , ΔK , ΔL , and ΔE-K , is not due to lower protein expression . Together , these data indicate that C-terminal region of MUT-16 ( J , K , and L ) contains a region essential for foci formation . In previous work we examined the requirements for Mutator foci formation [14] . In brief , MUT-16 is required for localization of MUT-2 , MUT-7 , RDE-2 , MUT-14 , and MUT-15 , all of which localize independently of one another except for MUT-7 , which requires RDE-2 for localization ( Fig 1C ) . RRF-1 and RDE-8 have both been previously shown to colocalize and co-immunoprecipitate with MUT-16 but whether they directly interact with MUT-16 or interact via other Mutator complex proteins was unclear [14–16] . We also suspected the Zc3h12a-like ribonucleases NYN-1 and NYN-2 to be Mutator complex proteins because we had identified them in an IP-mass spectrometry experiment with MUT-16 ( S1 Table ) . Additionally , Tsai et al . ( 2015 ) had identified them in a IP-mass spectrometry experiment with RDE-8 , as well as demonstrated that the nyn-1; nyn-2 double mutant displayed a marked reduction in WAGO-class 22G-RNAs , a phenotype similar to that of mutations in other Mutator complex components [16] . To determine whether NYN-1 localizes to the Mutator foci and to define the requirements for RRF-1 , RDE-8 , and NYN-1 localization , we generated GFP::RRF-1 , mCherry::RDE-8 , and mCherry::NYN-1 using CRISPR . Each protein forms distinct foci in germ cells and colocalizes with MUT-16 , indicating that NYN-1 is indeed a component of the Mutator foci ( S2A Fig ) . Tagged RRF-1 , RDE-8 , and NYN-1 strains were then crossed to strong loss-of-function mutations in other known Mutator foci components , including mut-16 , mut-2 , rde-2 , mut-14 smut-1 , and mut-15 . Of the genes tested , only mut-16 was required for RRF-1 localization ( S2B Fig ) . Localization of RDE-8 and NYN-1 to Mutator foci requires both mut-16 and mut-15 . To determine if NYN-1 or RDE-8 were required for each other’s localization we further crossed mCherry::NYN-1 to rde-8 mutants , and mCherry::RDE-8 to nyn-1; nyn-2 double mutants . mCherry::NYN-1 was able to localize independently of RDE-8 , whereas NYN-1 and NYN-2 were required for RDE-8 localization ( S2B Fig ) . To identify regions of the MUT-16 protein that are required for recruitment of other Mutator complex proteins we introduced the panel of mut-16 deletions into the MUT-2::GFP , mCherry::RDE-8 , mCherry::NYN-1 , GFP::RRF-1 , RDE-2::GFP , and MUT-14::GFP strains . Each of these six lines had a wild-type RNAi response in the presence of full-length mut-16 ( S3A Fig ) [14] . All deletion strains were generated independently of one another and deletions of the same region in different strain backgrounds behaved similarly to one another with respect to their effect on MUT-16 localization and ability to respond to germline and somatic RNAi ( S3B Fig ) . MUT-2::GFP foci were disrupted in the ΔB , ΔC , ΔJ , ΔK , and ΔL deletions ( Fig 4A–4D and S4 Fig ) , however MUT-16 localization was also disrupted in ΔJ , ΔK , and ΔL ( Figs 3A , 3D and 4D ) . To determine whether MUT-2 and MUT-16 could still interact at the molecular level in each deletion strain , we performed co-immunoprecipitation in each of the MUT-16 deletion backgrounds . The interaction between MUT-2 and MUT-16 was only disrupted in the ΔB and ΔC deletions , however , disruption of interactions between MUT-16 ΔC and other Mutator complex proteins could be at least partially due to the reduced expression of MUT-16 ΔC in the germline ( S5 Fig ) . Despite the reduction in visible Mutator foci in ΔJ , ΔK , and ΔL , MUT-2 still physically interacts with MUT-16 at levels similar to the full-length control when these deletions are present ( Fig 4E ) , indicating that ΔJ , ΔK , and ΔL regions are not required for recruitment of MUT-2 to the Mutator complex . This result suggests that when visible Mutator foci are disrupted in the ΔJ , ΔK , and ΔL mutants , many of the Mutator complex proteins may still interact in diffuse cytoplasmic complexes . We also observed that the MUT-16 protein is highly prone to degradation during the immunoprecipitation procedure . While full-length MUT-16 could be observed in most lanes , we also could detect multiple degradation products , with a prominent product of ~70kD in most lanes ( S5 Fig ) . We did not observe substantial degradation with other proteins we worked with and suspect the highly disordered nature of MUT-16 may leave it more exposed to proteases during the immunoprecipitation procedure . Similar to MUT-2 interaction , RDE-8 , NYN-1 , and MUT-14 depend on the ΔB and ΔC regions of MUT-16 for localization to the Mutator foci ( Fig 4A and S4 Fig ) . In contrast , RDE-2 fails to localize to Mutator foci in the ΔH-I deletion , and RRF-1 is at least partially disrupted in ΔB , ΔC , ΔF , and ΔG mutants ( Fig 4A and S4 Fig ) . Because RRF-1 foci are more difficult to detect than the other Mutator complex proteins , we also tested the physical interaction between MUT-16 and RRF-1 by co-immunoprecipitation in each of the MUT-16 deletion backgrounds . Only the ΔF mutant completely disrupted the interaction between MUT-16 and RRF-1 , though the amount of RRF-1 immunoprecipitated by the ΔB , ΔC , and ΔG mutants was modestly reduced relative to full-length MUT-16 ( Fig 4F ) . All together , these results suggest that different regions of MUT-16 are important for recruiting each of the Mutator complex proteins . These regions are separate and distinct from the C-terminal J , K , and L regions , which are important for Mutator foci formation . The J , K , and L regions of MUT-16 are each necessary for robust MUT-16 foci formation; to determine whether they are also sufficient for foci formation , we generated a series of C-terminal fragments of MUT-16 fused to GFP and inserted them into the genome using MosSCI ( Fig 5A and 5B ) [33] . Full-length MUT-16 forms foci throughout the germline with the brightest foci present in the mitotic region and the leptotene/zygotene regions of the germline ( Fig 5C ) [14] . To guarantee that any visible foci are not seeded by full-length , untagged MUT-16 , C-terminal MUT-16 fragments were introduced into a strain containing an early stop codon in the mut-16 gene prior to imaging . All C-terminal GFP fusion constructs containing the JKL region ( mut-16EFGHIJKL::gfp , mut-16GHIJKL::gfp , mut-16IJKL::gfp , and mut-16JKL::gfp ) had visible MUT-16 foci , but as the fusion proteins became smaller ( mut-16IJKL::gfp and mut-16JKL::gfp ) , the foci became less intense and fewer in number ( Fig 5D–5G ) . mut-16KL::gfp did not form foci , even in the presence of an endogenous , full-length copy of MUT-16 and despite robust expression of this MUT-16 fragment ( Fig 5B and 5H ) . These data are consistent with the JKL regions being necessary and sufficient for foci formation , but also indicate that additional regions of MUT-16 may help promote robust Mutator foci . We previously observed that Mutator foci form in the germline but not in somatic cells [15] . In the process of constructing an N-terminally GFP-3xFLAG-tagged MUT-16 , a strain construction intermediate was a mut-16 transcriptional reporter ( mut-16p::gfp ) at the endogenous mut-16 locus [34] . While imaging mut-16p::gfp , we observed that , while GFP fluorescence is present throughout the animal , the germline is distinctly brighter , suggesting that MUT-16 may be expressed at higher levels in the germline compared to the somatic tissues ( Fig 6A and 6B ) . Numerous studies have shown that phase separation is a concentration-dependent process [21–24 , 35] . We hypothesized that MUT-16 foci may form in germ cells but not in somatic cells because MUT-16 germ cell protein levels are above the concentration threshold at which MUT-16 phase separates to form foci . To test this hypothesis , we generated myo-3p::mut-16::gfp , which drives MUT-16 from a muscle-specific promoter and is expressed from a high-copy extrachromosomal array . This transgenic strain , in which the MUT-16 protein is overexpressed in muscle cells , has muscle-specific MUT-16::GFP foci ( Fig 6C , top row ) . The foci vary in size , some being substantially larger than endogenously expressed MUT-16 foci . The co-expressed mCherry protein ( myo-3::mCherry ) , which we also used as a marker for muscle cells , does not form foci but rather is expressed diffusely in both the cytoplasm and nucleus . Similarly , the GFP protein alone expressed under the myo-3 promoter ( myo-3::gfp ) and injected at the same molar concentration as myo-3p::mut-16::gfp , is expressed diffusely in the cytoplasm and nucleus of muscle cells ( Fig 6C , bottom row ) . A control strain which expressed only myo-3p::mCherry in the presence of mut-16::gfp expressed at the endogenous locus under its endogenous promoter ( mut-16p::mut-16::gfp ) did not have foci in muscle cells ( Fig 6C , middle row ) . Rather , diffuse cytoplasmic expression of endogenous mut-16p::mut-16::gfp is visible in the muscle and throughout the animal , indicating that over-expression of the mCherry protein in the muscle does not drive MUT-16 into ectopic foci . To determine if the size or quantity of ectopic MUT-16 foci change at different protein concentrations , we injected the myo-3p::mut-16::gfp plasmid at 20 ng/ul , 5 ng/ul , 1 ng/ul or 0 . 25 ng/ul . While there was variability between independent lines isolated from each set of injections , generally lines from injections at the highest concentration ( 20 ng/ul ) had larger and brighter foci than lines from lower concentrations ( 5 ng/ul , 1 ng/ul , or 0 . 25 ng/ul ) , whereas some lines from the lower concentration injections had no visible foci at all ( Fig 6D ) . In some very high expressing 20 ng/ul lines , we observed large condensates of MUT-16::GFP protein that are no longer spherical ( Fig 6D , top left ) . The presence of these larger structures could indicate that at very high concentrations MUT-16 is behaving more solid or gel-like , which has been seen previously for some IDR-containing proteins [24] . These data indicate that overexpression of MUT-16 in muscle cells is sufficient for somatic foci formation and suggest that MUT-16 can form foci in a concentration-dependent manner . To determine whether the ectopic foci formed by overexpression of MUT-16 in the muscle can recruit other Mutator proteins , we generated extrachromosomal arrays overexpressing myo-3p::mut-16::gfp in strains carrying MosSCI lines expressing either mut-15::mCherry or rde-2::mCherry . In a control strain expressing only myo-3p::mut-16::gfp; there is very little mCherry signal overlapping the ectopic MUT-16 foci ( Fig 6E , top row ) . In contrast , in the strains expressing mut-15::mCherry or rde-2::mCherry at endogenous levels , ectopic MUT-16 foci in the muscle recruit detectable levels of mut-15::mCherry or rde-2::mCherry ( Fig 6E , middle and bottom rows ) . Thus overexpression of MUT-16 is sufficient to drive not just MUT-16 , but also other Mutator complex proteins , into ectopic Mutator foci . Multiple studies have demonstrated that liquid-like condensates , but not solid aggregates , can be disrupted by aliphatic alcohols , such as 1 , 6-hexanediol [36–39] . These compounds disrupt weak , hydrophobic interactions , and have previously been shown to alter the permeability of the nuclear pore [40] . Addition of 5% 1 , 6-hexanediol to C . elegans gonads resulted in severely disrupted MUT-16 foci ( Fig 7A ) . Because a liquid-like state requires weak molecular interactions whereas tighter interactions promote a solid state [41] , these results support the hypothesis that Mutator foci have liquid-like properties . Phase-separated condensates can also be perturbed by changes in temperature . For example , the DEAD-box helicase Ddx4 , which is a component of germ granules , contains disordered regions that condense upon exposure to low temperatures and dissolve when returned to high temperatures [21] . Other disordered proteins behave in the opposite manner , condensing at high temperatures and dissolving at low temperatures [42] . To determine whether MUT-16 foci are temperature-sensitive , we subjected MUT-16::GFP animals to heat stress by placing them at 30°C for 6 hours . We returned the animals to room temperature ( ~21°C ) , and strikingly , the MUT-16 foci were no longer visible in the germ cells ( Fig 7B and S6 Fig ) . However , within 15 minutes , some animals have already reformed MUT-16 foci ( Fig 7B and S6 Fig ) . Over the full time course of 60 minutes at room temperature , many heat-shocked animals displayed no discernable differences in foci presence and intensity from animals raised at permissive temperatures ( Fig 7B , S6 Fig , and S1 Movie ) . These data indicate that MUT-16 foci can change in response to environmental conditions , such as temperature . Germ granule proteins , such as PGL-1 , PGL-3 , and LAF-1 in C . elegans P granules , recover rapidly after photobleaching indicating that the internal components of the granule can rearrange , similar to molecules in a liquid state [22 , 28 , 43 , 44] . To determine if MUT-16 foci are similarly liquid-like , we performed fluorescence recovery after photobleaching ( FRAP ) experiments . Due to the small size of Mutator foci ( <500 nm ) , we chose to bleach entire foci and measure the recovery of MUT-16 to the foci from the surrounding cytoplasm . MUT-16::GFP foci recovered rapidly after photobleaching , t1/2 = 7 . 2 ± 1 . 0 seconds ( SEM , n = 5 ) . Recovery only reaches ~35% of pre-bleached intensity ( Fig 7C and 7D and S1 Movie ) , indicating that there may be both a mobile fraction of MUT-16 that can exchange quickly between Mutator foci and the cytoplasm , and an immobile fraction that exchanges very slowly or not at all . Incomplete recovery of fluorescence has previously been observed for in vivo FRAP of P granules and of the adjacent and recently discovered Z granules [43–45] . Fluorescence recovery data , together with the dependence of Mutator foci on concentration , temperature , and hydrophobic interactions , suggest that Mutator foci have properties of a phase-separated condensate .
MUT-16 function can be subdivided across different regions of the protein ( Fig 7E ) . The largest structured region ( specifically the B-C region ) of MUT-16 recruits multiple proteins , including the nucleotidyl transferase MUT-2 , the DEAD-box RNA helicase MUT-14 , and the protein of unknown function MUT-15 , whose localization requirements we can infer based on the localization of the Zc3h12a-like ribonucleases RDE-8 and NYN-1 ( Fig 4A and 4C , and S2B Fig ) . MUT-2 , MUT-14 , and MUT-15 likely interact directly with MUT-16 , as we can robustly detect these interactions by immunoprecipitation and no other unknown proteins have been identified as part of this complex by IP-mass spectrometry experiments ( Fig 4 , S5 Fig and S1 Table ) [14] . Since MUT-15 recruits NYN-1 , NYN-2 and RDE-8 , at least six Mutator complex proteins localize to Mutator foci through the B-C region of MUT-16 ( S2B Fig and Fig 7E ) . It remains to be determined whether a single MUT-16 protein can interact with all of these proteins concurrently . Interestingly , while we initially hypothesized that the IDR would be important for promoting Mutator foci formation , we found regions of disorder also function to recruit proteins to the Mutator complex . In particular , region F recruits the RdRP protein RRF-1 and the H-I region recruits the exonuclease MUT-7 through its interaction with RDE-2 . The regions that directly recruit Mutator complex proteins correlate with regions required for robust RNAi in somatic tissues ( Fig 2 ) . Specifically , mut-16ΔC , ΔF , and ΔH-I have somatic RNAi defects nearly as severe as a mut-16 null allele . mut-16ΔL also has similarly severe somatic RNAi , however unlike ΔC , ΔF , and ΔH-I , deletion of the L region does not disrupt the interaction of MUT-16 with any other Mutator complex proteins of which we are aware . Since we cannot detect Mutator foci in the soma , we do not yet know how the loss of the L region affects Mutator complex formation or function in somatic cells . In addition , it is unclear why the remaining regions of MUT-16 are required for germline but not somatic RNAi . It is possible that they recruit yet uncharacterized germline-specific RNAi factors , or may facilitate currently unexamined interactions between primary and secondary Argonaute proteins and the Mutator complex . Overall , this data suggests that at least some of the most severe defects in somatic RNAi that occur after deleting portions of the mut-16 gene stem from loss of recruitment of specific proteins to the Mutator complex . Our fusion of C-terminal fragments of MUT-16 to GFP , demonstrates that a minimal region comprised of amino acids 773–1050 ( JKL region ) is sufficient for Mutator foci formation ( Fig 5 ) . This region is ~70% disordered , but contains the structured L region that is also key to the formation of Mutator foci . We did observe that including additional regions of the MUT-16 protein , specifically a larger portion of the disordered region , can promote increased size and number of Mutator foci . Interestingly , the regions with the highest glutamine and proline content are centered on the H-I and J regions of MUT-16 , only partially overlapping this region critical for Mutator foci formation . This finding suggests that isolated stretches of disordered amino acids are not sufficient to mediate phase separation , but require the neighboring protein environment to promote Mutator foci formation . Only recently has it become clear that many biological processes involve intracellular phase transitions , from spindle assembly to heterochromatin formation [37 , 42 , 46] . Proteins that are heterogeneous in conformation , such as proteins containing IDR , typically drive phase separation by providing a flexible platform for non-covalent interactions with nearby disordered proteins and/or ribonucleic acids . Other conditions such as protein or RNA concentration , post-translational modification , or changes in environmental conditions such as salt concentration or temperature can modulate these transitions [47] . Interestingly , we have been unable to co-immunoprecipitate MUT-16 with itself , which , while not conclusive , suggests that MUT-16 does not form strong intermolecular interactions and rather promotes Mutator foci formation via many weak or transient interactions . Based on our studies of MUT-16 , we propose that Mutator foci are phase-separated membrane-less organelles with liquid-like properties rather than a solid or aggregated structure , with five lines of evidence supporting this hypothesis . First , Mutator foci are roughly spherical in shape , as would be expected of a liquid compartment due to the cohesive forces of the surface layer ( i . e . surface tension ) . Second , Mutator foci dissolve in the aliphatic alcohol , 1 , 6-hexanediol , which disrupts only weak , hydrophobic interactions but not solids . Third , MUT-16::GFP foci recover rapidly , though incompletely , after photobleaching , indicating that a mobile fraction of MUT-16 can exchange freely between the Mutator foci and the cytoplasm . Fourth , Mutator foci only assemble when concentrations rise above a certain threshold , which is a hallmark of phase-separated condensates . And fifth , Mutator foci formation can be modulated by changes in temperature—increased entropy at higher temperatures can counteract phase separation and promote mixing of the condensed and bulk phases . The latter two points indicate that Mutator foci may be sensitive to intracellular and extracellular conditions , which could allow foci formation to be fine-tuned based on cell type or environmental conditions . A liquid-like nature of Mutator foci could allow for some proteins and RNA to freely exchange in and out of the compartment , while still maintaining a high concentration of factors required for small RNA amplification . Furthermore , the close juxtaposition of Mutator foci with nuclear pores and P granules , which also have liquid-like properties , suggests a model of RNA silencing where adjacent membrane-less organelles can exchange proteins and RNAs , but have intrinsic properties that make them immiscible with one another . A similarly close juxtaposition has been observed between P bodies and stress granules , and these types of interactions may reflect differences in surface tension of the two types of condensates [48 , 49] . Further characterization both in vivo and in vitro will be necessary to fully elucidate how Mutator foci form and interact with neighboring membrane-less organelles . Why are Mutator foci present in the germline but not the soma ? The answer may lie in the unique biophysical properties of MUT-16 , as well as in the concentration of and local environment surrounding MUT-16 in the germline . By concentrating Mutator complex proteins to the cytoplasmic side of nuclear pores , Mutator foci may be poised to capture deleterious RNAs during nascent RNA export . It is unclear why a similar mechanism is not also in place in somatic cells , but perhaps such a robust RNA silencing mechanism is not necessary in the soma as these cells will not be passed to the next generation . Mutator foci in the germline are exclusively associated with the nuclear periphery , until the germ cells become oocytes and the P granules detach from the nuclei and become cytoplasmic . Mutator foci and some nuclear pore components move with the P granules into the cytoplasm [14 , 43 , 50] . We do not yet know whether Mutator foci can interact directly with nuclear pores or if this interaction is mediated by the P granule or another unknown factor , however knockdown of P granule components does not disrupt the perinuclear localization of Mutator foci [14] . Interestingly , while some of the MUT-16 foci in myo-3p::mut-16::gfp are adjacent to the nuclear periphery , most are not ( Fig 6C ) . It is unclear whether the lack of perinuclear association in the muscle cells is due to the lack of a specific anchoring factor present only in germ cells , or perhaps the anchoring factor ( for example , nuclear pores ) is present but limited in supply relative to the substantial overexpression of MUT-16 . In conclusion , our results demonstrate that MUT-16 serves as a scaffold for assembly of Mutator foci and suggest that these foci form through liquid-liquid phase separation . The association of Mutator foci with P granules and nuclear pores suggests a model whereby newly transcribed mRNAs pass through nuclear pores into the P granule , where mRNAs are marked for silencing and targeted to the neighboring Mutator foci . Subsequently , siRNAs synthesized in the Mutator foci and perhaps secondary Argonaute proteins can transit back into the P granules to target additional complementary mRNAs for silencing . Thus , we propose that RNA silencing in germ cells may depend on RNA transit through phase-separated liquid compartments that concentrate RNA silencing enzymes .
The C . elegans wild-type strain is N2 . Worms were cultured at 20°C according to standard conditions unless otherwise stated [51] . Strains used include WM30 ( mut-2 ( ne298 ) I ) , NL3531 ( rde-2 ( pk1657 ) I ) , NL1810 ( mut-16 ( pk710 ) I ) , FX04844 ( nyn-2 ( tm4844 ) I ) , NL1820 ( mut-7 ( pk720 ) III ) , HT1593 ( unc-119 ( ed3 ) III ) , EG4322 ( ttTi5605 II; unc-119 ( ed9 ) III ) , GE24 ( pha-1 ( e2123 ) III ) , FX05004 ( nyn-1 ( tm5004 ) IV ) , FX02252 ( rde-8 ( tm2252 ) IV ) , GR1948 ( mut-14 ( mg464 ) smut-1 ( tm1301 ) V ) , and GR1747 ( mut-15 ( tm1358 ) V ) . All mutants were outcrossed at least 4x prior to any analysis . New strains made for this project are listed in S2 Table . All GFP and mCherry constructs were designed for insertion at the endogenous loci by CRISPR genome editing except the C-terminal fragments of mut-16 fused to GFP ( Fig 5 ) , which were integrated by Mos-mediated single-copy transgene insertion ( MosSCI ) [33] . For all CRISPR insertions of GFP or mCherry , we generated homologous repair templates using the primers listed in S3 Table . To create the mut-16::gfp::3xFLAG repair template , we first generated a GFP with C-terminal 3xFLAG tag and internal Floxed Cbr-unc-119 ( + ) , and inserted into the Kanamycin-resistant backbone of pDONR221 by isothermal assembly [52] . This cassette was then flanked with ~1 . 5-2kb of sequence from either side of the mut-16 stop codon , again by isothermal assembly . A similar design was used to create all mCherry repair templates . We first generated a 2xHA::mCherry[w/internal Floxed Cbr-unc-119 ( + ) ]::2xHA construct by cloning mCherry[w/internal Floxed Cbr-unc-119 ( + ) ] into XhoI/SpeI-digested pBluescript SK ( - ) . The mCherry[w/internal Floxed Cbr-unc-119 ( + ) ] was a gift from the lab of Jeremy Nance ( NYU ) . BamHI and BglII sites were engineered into the mCherry construct just following the ATG or just preceding the stop codon , respectively . This construct was sequentially digested with BglII , then BamHI; each digest was followed by ligation of a 2xHA oligo . This construct was then used as a PCR template to generate mCherry::2xHA[w/internal Floxed Cbr-unc-119 ( + ) ] or 2xHA::mCherry[w/internal Floxed Cbr-unc-119 ( + ) ] , and assembled with ~1kb of sequence from either side of the mut-16 stop codon , or the rde-8 and nyn-1 start codons , respectively . mut-2::gfp::3xFLAG , rde-2::gfp::3xFLAG , gfp::3xFLAG::rrf-1 , and gfp::3xFLAG::mut-16 ( for mut-16p::gfp ) repair templates were assembled into pDD282 ( GFP/3xFLAG with self-excising cassette , Addgene #66823 ) according to published protocols [34] . To protect the repair template from cleavage , we introduced silent mutations at the site of guide RNA targeting by incorporating these mutations into one of the homology arm primers or , if necessary , by performing site-directed mutagenesis [53] . The mut-16 guide RNA was cloned into PU6::unc-119_sgRNA ( Addgene #46169 ) by site-directed mutagenesis [54] . All other guide RNA plasmids were generated by ligating oligos containing the guide RNA sequence into BsaI-digested pRB1017 ( Addgene #59936 ) [55] . Guide RNA sequences are provided in S3 Table . C-terminal fragments of mut-16 fused to GFP were cloned into targeting vectors for MosSCI . The endogenous promoter and 3’UTR were amplified , along with either full-length coding sequence or C-terminal fragments of the coding sequence ( S3 Table ) . These amplicons were inserted along with GFP by isothermal assembly into SpeI-digested pCFJ151 ( Addgene #19330 ) [33 , 52] . A similar strategy was used to generate myo-3p::mut-16::gfp , except with the mut-16 promoter replaced by the myo-3 promoter . CRISPR injections were performed according to published protocols [34 , 53 , 56] . GFP/mCherry CRISPR injection mixes included 10–25 ng/μl repair template , 50 ng/μl guide RNA , 50 ng/μl eft-3p::cas9-SV40_NLS::tbb-2 3'UTR ( Addgene # 46168 ) , 2 . 5–10 ng/μl GFP or mCherry co-injection markers , and 10 ng/μl hsp-16 . 1::peel-1 negative selection ( pMA122 , Addgene #34873 ) . mut-16::gfp::3xFLAG and all mCherry constructs were injected into HT1593 ( unc-119 ( ed3 ) III ) . Floxed Cbr-unc-119 ( + ) cassettes were later excised using eft-3p::Cre ( pDD104 , Addgene #47551 ) [53] . mut-2::gfp::3xFLAG , rde-2::gfp::3xFLAG , and gfp::3xFLAG::rrf-1 were injected into the wild-type strain . mut-16 deletion injection mixes included 50 ng/μl oligo repair template , 25 ng/μl each of two mut-16 guide RNAs , 50 ng/μl pha-1 repair oligo and 50 ng/μl eft-3p:Cas9 + pha-1 guide RNA ( pJW1285 , Addgene #61252 ) . Mixtures were injected into pha-1 ( e2123 ) mutant animals already carrying mut-16::gfp::3xFLAG or mut-16::mCherry::2xHA and other fluorescently-tagged Mutator complex proteins and mut-16 deletions were identified by PCR according to pha-1 co-conversion protocol [56] . For MosSCI injections , we integrated transgenes into the ttTi5605 MosI site in strain EG4322 ( Ch . II ) following a published MosSCI protocol [33] . Injection mixes contained 50 ng/μl MosSCI-targeting vector , 50 ng/μl eft-3p::Mos1 transposase ( pCFJ601 , Addgene #34874 ) , 10 ng/μl rab-3p::mCherry ( pGH8 , Addgene #19359 ) , 2 . 5 ng/μl myo-2p::mCherry ( pCFJ90 , Addgene #19327 ) , 5 ng/μl myo-3p::mCherry ( pCFJ104 , Addgene #19328 ) , and 10 ng/μl hsp-16 . 1::peel-1 negative selection ( pMA122 , Addgene #34873 ) . Extra-chromosomal arrays were generated as follows: 20ng/ul myo-3p::mut-16::gfp , 5 ng/μl myo-3p::mCherry ( pCFJ104 ) and 70 ng/ul pBluescript injected into HT1593—unc-119 ( ed3 ) for cmpEx76 and cmpEx89; 5 ng/ul myo-3p::mut-16::gfp , 5 ng/μl myo-3p::mCherry ( pCFJ104 ) and 70 ng/ul pBluescript injected into HT1593—unc-119 ( ed3 ) for cmpEx90; 1 ng/ul myo-3p::mut-16::gfp , 5 ng/μl myo-3p::mCherry ( pCFJ104 ) and 70 ng/ul pBluescript injected into HT1593—unc-119 ( ed3 ) for cmpEx91; 0 . 25 ng/ul myo-3p::mut-16::gfp , 5 ng/μl myo-3p::mCherry ( pCFJ104 ) and 70 ng/ul pBluescript injected into HT1593—unc-119 ( ed3 ) for cmpEx92; 5 ng/μl myo-3p::mCherry ( pCFJ104 ) and 70 ng/ul pBluescript injected into the mut-16 ( cmp3[mut-16::gfp::3xFLAG] ) strain for cmpEx79; 7 . 5 ng/ul myo-3p::gfp ( pPD118 . 20 ) and 70 ng/ul pBluescript into the wild-type strain for cmpEx88; and 20ng/ul myo-3p::mut-16::gfp and 70 ng/ul pBluescript injected into HT1593—unc-119 ( ed3 ) , mut-15::mCherry , and rde-2::mCherry strains for cmpEx76 , cmpEx86 , and cmpEx87 . C . elegans were imaged in M9 buffer with sodium azide to prevent movement . In some cases , to obtain higher quality images , animals were dissected prior to mounting . For scoring presence or absence of foci , deletions were blinded and scored by multiple individuals . A minimum of 10 animals were scored for each strain . Imaging was performed on a DeltaVision Elite microscope ( GE Healthcare ) using a 60x N . A . 1 . 42 oil-immersion objective . When data stacks were collected , three-dimensional images are presented as maximum intensity projections . Images were pseudocolored using the SoftWoRx package or Adobe Photoshop . For quantification of somatic and germline fluorescence intensity in the mut-16p::gfp strain , 31 age-matched L4 animals were imaged using identical microscope settings . The somatic and germline tissue was traced in ImageJ and mean gray value was calculated for each region . 1 , 6-hexanediol treatment was performed by dissecting adult MUT-16::GFP animals in M9 buffer with or without the addition of 5% 1 , 6-hexanediol and imaged immediately following dissection . For heat stress experiment , plates of L4 stage MUT-16::GFP animals were wrapped in Parafilm and placed in a 30°C incubator for 6 hours . We then removed the plates from the incubator , immediately transferred the animals to slides , and began imaging . No more than 5 minutes elapsed between removal from 30°C and collection of the first time point ( Fig 7B , 0 min ) . Images were captured every 3 minutes for 60 minutes at room temperature ( ~21°C ) . Fluorescence recovery after photobleaching ( FRAP ) was performed on a Leica SP8 Falcon laser scanning confocal microscope using a 63x N . A . 1 . 4 oil-immersion objective . Two images were acquired prior to bleaching , followed by bleaching , then 10 images at 0 . 44 second intervals and 10 additional images at 10 second intervals . Data was analyzed using the Leica Application Suite X software . Immunoprecipitations were performed as previously described [14] . ~40 , 000 synchronized adult C . elegans ( ~68 h at 20°C after L1 arrest ) were collected , frozen in liquid nitrogen , and homogenized using a mortar and pestle . After further dilution into lysis buffer ( 1:10 packed worms:buffer ) , insoluble particulate was removed by centrifugation and a sample was taken as “input . ” The remaining lysate was used for the immunoprecipitation . HA-tagged proteins were immunoprecipitated using anti-HA affinity matrix ( ThermoFisher 26181 ) . For Western blots , proteins were resolved on 4–12% Bis-Tris polyacrylamide gels ( ThermoFisher ) , transferred to nitrocellulose membranes , and probed with anti-HA 1:1 , 000 ( Roche 12013819001 ) , anti-FLAG 1:1 , 000 [M2 clone] ( Sigma-Aldrich F1804 ) , anti-actin 1:10 , 000 ( Abcam ab3280 ) , or anti-GFP 1:100 ( Riken BRC JFP-J5 ) [57] . Secondary HRP antibodies were purchased from ThermoFisher . We observed that MUT-16 was often substantially degraded during the immunoprecipitation procedure . We attempted to remedy this problem by reducing the time samples spent in lysis buffer , shortening the immunoprecipitation step , and increase the amount of protease inhibitor . These steps resulted in little improvement of full-length MUT-16 yield in both input and IP samples . We also noted that the MUT-16 degradation product was present in full-length and all MUT-16 deletion mutants except for ΔH-I . Given that MUT-16 is tagged at the C-terminus , the deletions preceding ΔH-I all form degradation products of the same size , and the deletions following ΔH-I all vary in size in proportion to the relative deletion sizes , we suspect that the degradation site in MUT-16 is somewhere in the ΔH-I region ( S5 Fig ) . For mass spectrometry experiments , immunoprecipitation was performed as described above , starting with ~500 , 000 synchronized adult C . elegans ( ~68 h at 20°C after L1 arrest ) for each sample . GFP and FLAG immunoprecipitation was performed using anti-GFP affinity matrix [RQ2 clone] ( MBL International D153-8 ) and anti-FLAG affinity matrix [M2 clone] ( Sigma-Aldrich A2220 ) . After immunoprecipitation , samples were precipitated using the ProteoExtract Protein Precipitation Kit ( EMD Millipore 539180 ) and submitted to the Taplin Mass Spectrometry facility at Harvard Medical School for protein identification . For RNAi assays , L1 animals were fed E . coli expressing dsRNA against pos-1 , lin-29 , or nhr-23 [58] . For pos-1 , animals were scored 4 days later for hatching of the F1 embryos . For lin-29 or nhr-23 , animals were scored 2–3 d later for vulval bursting or larval arrest , respectively . To quantify intermediate phenotypes of nhr-23 and lin-29 RNAi , we established two subcategories to distinguish between partial RNAi defects ( animals are smaller than wild-type with few or no eggs as adults , and for lin-19 , often have a protruding vulva ) and total RNAi defects ( animals are phenotypically wild-type ) . | Small RNAs are a driving force behind the regulation of both essential genes and deleterious transcripts . The Mutator complex is critical to the amplification of high levels of small RNAs and it requires the protein MUT-16 for its assembly . Here we investigate the function of MUT-16 by generating small deletions in the mut-16 gene . Through analysis of the subsequently altered protein , we demonstrate that MUT-16 functions as a scaffold , bringing together many other proteins required for small RNA biogenesis and amplification . Furthermore , we identified a fragment of MUT-16 that is sufficient to promote assembly of MUT-16 into foci that are dynamic and responsive to environmental conditions . We propose that these Mutator foci behave like liquid droplets within the cell , similar to the immiscibility of oil droplets in water . Mutator foci localize to the periphery of germ cell nuclei near P granules , which also have liquid-like properties and contain many factors involved in RNA silencing . Thus , our data suggest that RNA silencing is mediated by compartments of RNAs and proteins in liquid-like assemblies at the periphery of germ cell nuclei . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"invertebrates",
"deletion",
"mutation",
"rna",
"interference",
"caenorhabditis",
"gene",
"regulation",
"animals",
"animal",
"models",
"germ",
"cells",
"mutation",
"caenorhabditis",
"elegans",
"model",
"organisms",
"immunoprecipitation",
"experimental",
"organism",
"systems",
"silent",
"mutation",
"epigenetics",
"cell",
"nucleus",
"cellular",
"structures",
"and",
"organelles",
"nuclear",
"pores",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"genetic",
"interference",
"animal",
"cells",
"gene",
"expression",
"precipitation",
"techniques",
"biochemistry",
"rna",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"non-coding",
"rna",
"organisms"
] | 2018 | Distinct regions of the intrinsically disordered protein MUT-16 mediate assembly of a small RNA amplification complex and promote phase separation of Mutator foci |
Differential DNA methylation defects of H19/IGF2 are associated with congenital growth disorders characterized by opposite clinical pictures . Due to structural differences between human and mouse , the mechanisms by which mutations of the H19/IGF2 Imprinting Control region ( IC1 ) result in these diseases are undefined . To address this issue , we previously generated a mouse line carrying a humanized IC1 ( hIC1 ) and now replaced the wildtype with a mutant IC1 identified in the overgrowth-associated Beckwith-Wiedemann syndrome . The new humanized mouse line shows pre/post-natal overgrowth on maternal transmission and pre/post-natal undergrowth on paternal transmission of the mutation . The mutant hIC1 acquires abnormal methylation during development causing opposite H19/Igf2 imprinting defects on maternal and paternal chromosomes . Differential and possibly mosaic Igf2 expression and imprinting is associated with asymmetric growth of bilateral organs . Furthermore , tissue-specific imprinting defects result in deficient liver- and placenta-derived Igf2 on paternal transmission and excessive Igf2 in peripheral tissues on maternal transmission , providing a possible molecular explanation for imprinting-associated and phenotypically contrasting growth disorders .
Imprinted genes show monoallelic and parent-of-origin-dependent expression and play key roles in the control of growth and development . In humans , altered expression of imprinted genes is associated with Imprinting Disorders ( IDs ) that are characterized by growth , metabolic and behavioural disturbances [1–2] . Most imprinted genes are organized in clusters , in which their parental-specific expression is dependent on Imprinting Control Regions ( ICRs ) . ICRs correspond to 2–5 kb-long sequences with differential DNA methylation on their maternal and paternal alleles . Parental-specific ICR methylation is acquired during gametogenesis and maintained in the zygote and somatic cells throughout development despite extensive demethylation occurring in the embryo before implantation and de novo methylation after implantation [3] . An evolutionary conserved cluster of imprinted genes of about 1 Mbp is located on human chromosome 11p15 . 5 and mouse distal chromosome 7 . The cluster is organized in two functionally independent domains , each with its own ICR . In the telomeric domain , the H19/IGF2 intergenic differentially methylated region ( also known as and herein termed Imprinting Center 1 , IC1 ) controls the reciprocal imprinting of the maternally expressed H19 and paternally expressed Insulin like Growth Factor 2 ( IGF2 ) genes [3] . IGF2 is required for normal foetal growth [4] . The liver is the main endocrine source of IGF2 in post-natal life , but autocrine/paracrine activity is found in most embryonic tissues , particularly in placenta , where it is needed for correct allocation of maternal resources to fetal growth [5 , 6] . H19 is a long non-coding RNA with inhibitory activity on foetal growth [7] . Both IGF2/Igf2 and H19 are down-regulated after birth in both humans and mice but their deficiencies have long-lasting effects on somatic growth [4 , 7–11] . In mouse embryos , H19 and Igf2 are co-expressed in endoderm- and mesoderm-derived tissues , and their expression depends on the same downstream enhancers on the maternal and paternal chromosomes , respectively [12–15] . IC1 is structurally different in humans and mice—human IC1 ( hIC1 ) is ~ 5kb-long and contains seven CCCTC-binding factor ( CTCF ) target sites ( CTS ) , whereas mouse IC1 ( mIC1 ) is ~ 2kb-long and contains four CTS ( Fig 1A ) . CTCF binding to IC1 is required for the formation of an insulator with enhancer blocking activity in both species [16–17] . Because IC1 is methylated on the paternal allele and CTCF binding is inhibited by DNA methylation , the insulator is formed only on the maternal chromosome where it prevents the activation of IGF2 but allows activation of H19 by the enhancers . The opposite happens on the paternal chromosome , where IGF2 is activated and H19 silenced . Opposite hIC1 methylation and imprinting defects are associated with the Beckwith-Wiedemann syndrome ( BWS , MIM #130650 ) and the Silver Russell syndrome ( SRS , MIM #180860 ) , two IDs characterised by congenital overgrowth and congenital undergrowth , respectively [1] . In particular , hIC1 gain of methylation ( GOM ) resulting in IGF2 activation and H19 repression on the maternal chromosome is found in 5–10% of BWS cases . Conversely , hIC1 loss of methylation ( LOM ) leading to IGF2 repression and H19 activation on the paternal chromosome occurs in about 50% of SRS patients . In a number of BWS cases with hIC1 GOM , small deletions and single nucleotide variations within hIC1 co-segregate with the clinical phenotype and abnormal methylation upon maternal transmission but lead to normal phenotype on paternal transmission [11 , 18–25] . Different paternally inherited hIC1 deletions have been recently described in a few cases of SRS with IC1 LOM [26] . Although the extent of these deletions is similar to those found in BWS families , the sequence and the CTSs involved are different . In the last twenty years , several mutations have been introduced into the endogenous mIC1 locus [27–35] . This work has been instrumental to demonstrate the fundamental role that H19/IGF2 imprinting has in the aetiology of congenital overgrowth and undergrowth associated with imprinting defects . However , due to its structural differences with the orthologous mouse locus , the mechanism by which hIC1 mutations affect epigenotype and phenotype in both BWS and SRS is still obscure . To clarify the role of hIC1 mutations in the origin of imprinting defects and in the pathogenesis of BWS and SRS , we generated a knock-in ( KI ) mouse line in which the endogenous mIC1 was replaced by the orthologous hIC1 allele carrying a mutation ( hIC1Δ2 . 2 ) that is associated with BWS on maternal transmission [36] . We compared the H19hIC1Δ2 . 2 line with wildtype mice and the previously described line carrying a humanized H19 allele with the wildtype human ICR ( H19hIC1 ) [37] . The results demonstrate growth and molecular abnormalities of the mice with maternal and paternal transmission of the mutant KI that resemble those of BWS and SRS , respectively , including asymmetric organ growth . Importantly , tissue-specific and mosaic dysregulation of H19/Igf2 imprinting indicates new pathogenetic mechanisms of congenital growth disorders and lateralized/regional over/under-growth associated with imprinting defects .
In order to study the relationship between genotype , epigenotype and phenotype of IDs in the mouse , we replaced the endogenous mIC1 with a mutant hIC1 allele ( H19hIC1Δ2 . 2 ) previously found in BWS [19 , 36] , by homologous recombination in mouse embryonic stem ( ES ) cells ( Fig 1A ) . Chimeras were obtained and germ line transmission was confirmed by Southern blotting ( Fig 1B ) . The transgenic line was then bred to pCX-NLS- Cre transgenic line to remove the NeoR cassette and its excision was confirmed by PCR ( Fig 1C . See also Materials and methods ) . We crossed KI mice on a C56BL/6 ( B6 ) background to Balb/C and used polymorphisms present between these two strains to distinguish parental alleles . To compare the behaviour of the mutant with that of the wildtype hIC1 allele ( H19hIC1 ) in the same genetic background , the previously described H19hIC1 line [37] was generated anew by targeting ES cells and breeding the mice using similar procedures to what done with the H19hIC1Δ2 . 2 line . Subsequent experiments were performed on both humanized KI lines . The wildtype +/+ littermates were also assayed as control . To investigate the presence of kidney asymmetry , a clinical sign often found in BWS [2] , we measured the weight of left and right kidneys of the mice carrying the mutant KI . Maternal transmission was first assessed . A significant difference between the two kidneys ( with no bias toward the left or right organ ) was found in adult and newborn H19hIC1Δ2 . 2/+ mice , but not in H19+/+ littermates ( Fig 8A and 8B ) . In contrast , no difference was observed in H19hIC1/+ mice ( S7 Fig ) . hIC1 methylation and H19/Igf2 expression were then assessed in the organs of the neonates carrying the mutant KI . While DNA methylation and H19 expression were comparable ( Fig 8C and 8D ) , significant differences of total and allele-specific expression of Igf2 were found between the heavier and lighter kidneys of H19hIC1Δ2 . 2/+ ( with higher expression in the larger organ ) mice ( Fig 8E and 8F ) . Next , we investigated the presence of kidney asymmetry in the mice with paternal transmission of the KI . As for maternal transmission , weight differences between the two kidneys in H19+/hIC1Δ2 . 2 mice were significantly higher than in their H19+/+ littermates , both at neonatal and adult stages ( Fig 8G and 8H ) . Also , comparable hIC1 methylation and global H19 RNA levels were found in left and right kidneys of all tested animals ( Fig 8I and 8J ) . In contrast , paternal H19 expression was relatively more up-regulated in the lighter kidneys of the H19+/hIC1Δ2 . 2 mice ( Fig 8K ) . Although not statistically significant ( P<0 . 1 ) , a trend toward stronger Igf2 repression was observed in the smaller with respect to the larger kidney ( Fig 8I ) . Having demonstrated that H19hIC1Δ2 . 2 is methylated in somatic cells upon both maternal and paternal transmission , we asked if methylation was already present in germ cells . For this purpose , we measured DNA methylation levels in oocytes and sperm by pyrosequencing . As previously demonstrated , H19hIC1 remains properly hypomethylated in female gametes but methylation is inefficiently established on H19hIC1 in male gametes ( Fig 9A and Ref . 37 ) . Similarly , H19hIC1Δ2 . 2 methylation was close to 0% and comparable with the endogenous mIC1 in oocytes ( Fig 9A ) . mIC2 was also analysed as control and to rule out contamination of somatic cells . The expected methylation value close to 100% was found in H19hIC1Δ2 . 2/+ , H19hIC1/+ and H19+/+ oocytes . In sperm , methylation levels were relatively low ( 10–14% ) on H19hIC1Δ2 . 2 as well as H19hIC1 ( 7–22% ) , while the endogenous mIC1 was almost 100% methylated ( Fig 9A ) . The endogenous mIC2 , analysed as control , was correctly unmethylated in all three lines ( H19hIC1Δ2 . 2/+ , H19hIC1/+ and H19+/+ ) . The low methylation status of H19hIC1Δ2 . 2 was confirmed by bisulphite sequencing , in both oocytes and sperm ( Fig 9B ) . Overall , these results demonstrate that while DNA methylation of mutant hIC1 is normally absent in oocytes , methylation is not efficiently established on the mutant KI as well as the wildtype KI in male germ cells , indicating that paternal hypomethylation of the H19hIC1Δ2 . 2 allele is seemingly acquired as early as germline development and persists into embryo development , while maternal hypermethylation of hIC1Δ2 . 2 alleles does not occur until after fertilization .
Gain and loss of IC1 methylation result in H19/IGF2 imprinting defects that are characteristic of BWS and SRS , respectively . In some patients , genetic mutations have been found associated with DNA methylation abnormalities in cis . However , the mechanism by which the genotype affects the epigenotype and phenotype in these cases is unknown . By employing a KI mouse line , this study demonstrates that a human genetic IC1 mutation reproduces several molecular and phenotypic abnormalities of BWS and SRS . The analysis of this mouse model provides mechanistic insights into the origin of prenatal overgrowth and undergrowth associated with H19/Igf2 imprinting defects , which are useful for understanding the aetiopathogenesis of BWS and SRS . We have previously demonstrated that maternal transmission of hIC1 can functionally replace mIC1 in the mouse by properly regulating IC1 methylation and H19/Igf2 imprinting [37] . We now observe that mice maternally inheriting the mutant KI allele acquire methylation on hIC1Δ2 . 2 after fertilization , exhibit H19 repression and biallelic Igf2 activation , and pre- and post-natal overgrowth . Upon paternal transmission , hIC1 lacks methylation resulting in complete Igf2 silencing , H19 activation , severe growth restriction particularly in placenta and embryonic lethality [37] . In contrast , paternal hIC1Δ2 . 2 is partially methylated and results in a more moderate imprinting defect and pre/post-natal undergrowth which is compatible with life . Methylation is established only in a minority of male germ cells on both hIC1Δ2 . 2 and hIC1 and appears unstable indicating evolutionarily divergent mechanisms of imprinting establishment between human and mouse [37] . However , while hIC1 is completely unmethylated , hIC1Δ2 . 2 shows partial methylation on both paternal and maternal chromosomes in embryo and placenta . Thus , methylation is likely acquired de novo and in mosaic form on hIC1Δ2 . 2 in somatic cells , during development . The functional difference between hIC1Δ2 . 2 and hIC1 is likely due to the lack of three CTSs in the mutant allele , which shows lower affinity for CTCF in human cells [36] . The results are consistent with hIC1 having an intrinsic propensity to acquire methylation that is inhibited by CTCF binding [32] . The different behaviour of hIC1Δ2 . 2 and mIC1 , which share a similar number of CTSs ( 4 ) and ZFP57 binding sites ( 6 ) , suggests that the CTSs alone are not sufficient to maintain insulator function and that CTS spacing or other transcription factor binding sites contribute to IC1 function . It is possible that one or more of these elements are reduced or missing and this exposes the mutant hIC1 allele to the action of de novo DNA methyl-transferases in pre- and/or post-implantation embryos . The mechanisms by which hIC1 methylation is altered in BWS and SRS are unknown . Our mouse model indicates that the maternal hIC1Δ2 . 2 methylation is acquired post-zygotically in BWS , but does not allow to distinguish if the partial IC1 methylation of SRS is due to a primary germ cell imprint establishment defect , or a post-zygotic maintenance defect , or both . However , the observation that the hIC1 methylation status can drastically change from gametes to somatic cells , suggests that maintenance mechanisms have a critical role in the origin of imprinting defects on both maternal and paternal chromosomes . Molecular analyses show that , although hIC1Δ2 . 2 is similarly methylated in neonatal liver , kidney and tongue , allele-specific H19/Igf2 expression is differently altered , suggesting that hIC1 enhancer blocking function is regulated in a tissue-specific manner ( Fig 10 ) , consistent with what was previously shown for mIC1 [35] . In particular , the insulator activity of hIC1Δ2 . 2 appears to be robust in liver and placenta , as demonstrated by weak expression of the Igf2 allele in cis with the KI . In contrast , the relatively high Igf2 expression indicates that the insulator activity of hIC1Δ2 . 2 is weak in tongue . The kidney shows intermediate and mosaic insulator activity resulting in differential Igf2 expression between left and right organs ( see below ) . Tissue-specific differences in insulator activity may result from different post-translational modifications of CTCF [35] . Importantly , the observation that Igf2 expression is properly regulated in specific tissue contexts in the presence of abnormal IC1 methylation paves the way to new exciting avenues for BWS and SRS therapy . Concerning H19 , this gene is significantly down-regulated in all three tissues of H19hIC1Δ2 . 2/+ mice and up-regulated in liver and tongue of H19+/hIC1Δ2 . 2 mice , with respect to wildtype littermates , consistent with the partial methylation of hIC1Δ2 . 2 and the demonstrated repressor activity of methylated IC1 [32] . In liver of H19+/hIC1Δ2 . 2 neonates , a strong derepression of the imprinted allele was not accompanied by a global ( mat + pat ) increase of the H19 RNA , possibly because of other physiological perturbations limiting its expression . Overall , gene expression results are consistent with the hypothesis that the organomegaly of H19hIC1Δ2 . 2/+ mice is associated with autocrine/paracrine effects of IGF2 , while defective hepatic and placental IGF2 expression underlie the growth restriction of H19+/hIC1Δ2 . 2 mice . Opposite deregulation of the growth inhibitory H19 transcript is likely playing an additional role in the growth abnormalities of the mice with maternal and paternal transmission of the mutant KI . Although in humans this mutation is associated only with BWS , the hIC1Δ2 . 2 mice display several features of BWS and SRS , on maternal and paternal transmission of the KI , respectively . Growth abnormalities originate prenatally and persist in adulthood . In particular , H19+/hIC1Δ2 . 2 mice do not catch-up growth during development , as seen in the majority of children with SRS [38] . In addition , nephromegaly and macroglossia that are observed in H19hIC1Δ2 . 2/+ mice are also distinctive clinical signs of BWS with IC1 molecular defects [39] . Finally , the tissue-specific differences of hIC1 function we observed in the mouse may explain some of the aetiopathogenetic mechanisms of BWS and SRS . The model we propose predicts that maternal IC1 GOM in BWS causes IGF2 activation primarily in peripheral tissues , such as tongue and kidney resulting in macroglossia and organomegaly , while paternal IC1 LOM in SRS causes IGF2 repression primarily in liver and placenta leading to deficient growth stimulation through defective endocrine secretion and placenta function . Both H19hIC1Δ2 . 2/+ and H19+/hIC1Δ2 . 2 mice show differential growth of the kidneys indicating that asymmetric growth of bilateral organs occurs in these animals , as in many cases of BWS with IC1 defects [40] . Derepression of the maternal Igf2 allele and repression of the paternal Igf2 allele are incomplete in the kidney of P1 H19hIC1Δ2 . 2/+ and P1 H19+/hIC1Δ2 . 2 mice , respectively ( see Figs 3 and 6 ) , indicating mosaic expression of H19/Igf2 in these tissues . Further analyses showed that , although IC1 methylation and H19 RNA levels were similar , Igf2 expression was higher in the larger than in the smaller kidney due to imprinting defects , indicating that Igf2 is mostly responsible for the asymmetric kidney growth . These results raise the hypothesis that mosaic IGF2 expression may also cause lateralized somatic overgrowth in humans . Our results are consistent with the findings of Ginart et al [41] demonstrating that the incomplete derepression of the paternal H19 allele in mutant mice can result from an epigenetic mosaicism at a single cell level . Overall , our findings show that a mutant human IC1 sequence can reproduce the opposite growth and molecular phenotypes of BWS and SRS in mouse , when introduced at the orthologous locus . Several mouse lines carrying mutations in the H19/Igf2 locus have been described so far . Some of these mice , including a 1 . 3 kb deletion of mIC1 ( Δ2 , 3 ) that results in tissue-specific loss of Igf2 imprinting , show similarities with our H19hIC1Δ2 . 2 model [28 , 34–35] . However , the methylation defects , growth abnormalities and H19/Igf2 dysregulation of H19hIC1Δ2 . 2 more closely reproduce the phenotypic features and contribute better in understanding the molecular pathogenesis of BWS and SRS . Such humanized mouse models will be useful for more accurately unravelling pathogenetic mechanisms and for developing new therapeutic strategies in these rare congenital growth disorders .
To generate the H19hIC Δ2 . 2 mouse line , we performed gene targeting by homologous recombination in E14 embryonic stem ( ES ) cells [42] to target the endogenous mIC1 with a plasmid containing the H19hIC1Δ2 . 2 allele and neomycin resistance cassette ( NeoR; Fig 1A ) . Briefly , a PciI–MluI restriction fragment of 620 bp spanning the break-point of the 2 . 2 kb deletion found in a BWS family [18] was extracted from the EΔ2 . 2 ( B5/b1 ) pL vector [36] and subcloned in the Δ3 . 8kb-5’ pre-targeting vector containing the wildtype hIC1 region [28] . Sanger sequencing of the fragment was performed and no variant was found in respect to the reference human genome hg19 . The subsequent steps to obtain the targeting vector were performed as previously described [37] . To compare H19hIC1 and H19hIC1Δ2 . 2 in the same strain background and avoid animal transfer from US to Italy , the hIC1 KI line was generated anew by performing gene targeting in E14 ES cells with the original vector used in the previous study [37] . Injection into B6 blastocysts of the H19hIC1-neo and H19hIC1Δ2 . 2-neo KI ES clones and generation of chimeras were performed by Cogentech Facility S . c . a . r . l . ( Milan , Italy ) . Chimeras were crossed to B6 mice and germline transmission of the KI was confirmed in the agouti progeny by PCR-genotyping using primers flanking the deletion break point ( hDMDB5SeqF: 5’–GGTAGTGAGGGATAGAACAC– 3’; hDMDB1RepR: 5’–GAGTGTCCTATTCCCAGATGAC– 3’ ) ( Fig 1B ) . The NeoR cassette was excised by crossing heterozygous KI with pCX-NLS- Cre transgenic mice [43] on a B6 background . Excision was tested by PCR using primers flanking the NeoR cassette ( NeoEXL3: 5’–ACAGAATCGGTTGTGGCTGT– 3’ H19SeqR1: 5’–CCACAGAGTCAGCATCCAC– 3’ ) ( Fig 1C ) . KI mice without the NeoR cassette were crossed with B6 mice and only the progeny carrying the KI and lacking the Cre gene were selected to expand and assay the KI lines . All animal experimentation was conducted in accordance with the guidelines of the Animal Care and Use Committee of Campania University “Luigi Vanvitelli” ( Naples , Italy ) and was authorized by the Italian Ministry of Health . Liver , kidneys and tongue were collected from mice at birth and at 14 weeks of age . Placenta and whole body excluding the head were recovered from conceptuses of 15 . 5 days post coitum ( E15 . 5 ) . Genomic DNA was isolated from tissues following the standard protocol of proteinase K digestion and phenol-chloroform extraction . Total RNA was extracted using TRI Reagent ( Sigma-Aldrich Italia , Milan ) and following the manufacturer’s protocol . Concentrations of nucleic acids were determined with Nanodrop spectrophotometer . Sperm was isolated from adult mice and DNA was extracted as described previously [44] . Unfertilized oocytes were collected from 4–5 superovulated females of 8 weeks and resuspended in 0 . 03% SDS , 10 mg glycogen , 10 mg proteinase K and 1 x PBS to a final volume of 20μl . Suspension was incubated at 37°C for 90 min followed by 15 min at 95°C before sodium bisulphite treatment for DNA methylation analysis ( see below ) . About 1 μg of total RNA was treated with RNase-free DNase , and first-strand cDNA was synthesized using the QuantiTech Reverse Transcription Kit ( Qiagen Italy , Milan ) , according to the manufacturer’s protocol . Total expression of H19 and Igf2 was measured by SYBR Green quantitative real-time RT-PCR ( Applied Biosystems Italy , Milan ) . Reactions were set up in triplicate and run on ABI PRISM 7500 using the default cycling conditions . Relative expression was determined using the ΔΔCT method , and the gene expression values were normalized to the expression of the Gapdh , Arpp0 and beta-actin reference genes . Primer sequences are available on request . Allele-specific expression analysis was performed by typing for the polymorphisms present in the F1 progeny between the C57BL/6 ( B6 ) and Balb/C mouse strains . The MspI Restriction Fragment Length Polymorphism ( RFLP ) of H19 ( GRCm38/mm10 chr 7:142 , 577 , 609; sequenced region: chr7:142 , 577 , 530–142 , 577 , 732 ) and the ( CA ) n repeat of Igf2 ( GRCm38/mm10 chr 7:142 , 652 , 936–142 , 652 , 973; sequenced region: chr7:142 , 652 , 821–142 , 653 , 091 ) were analysed as described by Pedone et al , [45] . The forward primer of Igf2 was labelled with FAM and the PCR products were run on the ABI 3130XL fluorescent capillary system ( Applied Biosystems Italy , Milan ) . The methylation status of cytosines in gDNA was determined by bisulphite treatment followed by pyrosequencing or cloning and sequencing . About 1 μg of genomic DNA was treated by sodium bisulphite using the Epitech Kit ( Qiagen Italy , Milan ) , according to the manufacturer’s instructions . For the pyrosequencing , converted DNA was amplified with primers of which the reverse primer was biotinylated . The PCR products were run on the PyroMark Q24 platform , using PyromMark Gold Q96 Reagent [Qiagen Italy , Milan] . For bisulphite cloning and sequencing , converted DNA was amplified , the PCR products were cloned in Topo pCR2 . 1 vector ( Topo-TA cloning kit , Termo Fisher Scientific Italy , Milan ) and the clones were sequenced by Sanger method at Microtech Sequencing Core ( Naples , Italy ) . Primers and PCR conditions are reported in S1 Table . Unless otherwise indicated , data are expressed as the mean ± standard error of the mean ( SEM ) . The significance of the difference between two groups ( KI and wildtype ) was determined with a two-tailed Student’s t-test with two-sample unequal variance . The number of samples , animals , biological or technical replicates are indicated in the respective figure legends . Differences with P-values ≤ 0 . 05 were considered significant . | A humanized mouse line carrying a mutation of the H19/IGF2 imprinting control region demonstrates how tissue-specific and mosaic imprinting alterations result in growth disorders with opposite clinical pictures and asymmetric growth of bilateral organs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"disorders",
"of",
"imprinting",
"animal",
"models",
"developmental",
"biology",
"model",
"organisms",
"methylation",
"tongue",
"experimental",
"organism",
"systems",
"beckwith-wiedemann",
"syndrome",
"epigenetics",
"dna",
"embryos",
"kidneys",
"dna",
"methylation",
"chromatin",
"digestive",
"system",
"research",
"and",
"analysis",
"methods",
"embryology",
"chromosome",
"biology",
"gene",
"expression",
"chemistry",
"chromatin",
"modification",
"mouse",
"models",
"dna",
"modification",
"clinical",
"genetics",
"biochemistry",
"mouth",
"cell",
"biology",
"nucleic",
"acids",
"anatomy",
"genetics",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"renal",
"system",
"physical",
"sciences"
] | 2018 | Tissue-specific and mosaic imprinting defects underlie opposite congenital growth disorders in mice |
In vertebrate definitive hematopoiesis , nascent hematopoietic stem/progenitor cells ( HSPCs ) migrate to and reside in proliferative hematopoietic microenvironment for transitory expansion . In this process , well-established DNA damage response pathways are vital to resolve the replication stress , which is deleterious for genome stability and cell survival . However , the detailed mechanism on the response and repair of the replication stress-induced DNA damage during hematopoietic progenitor expansion remains elusive . Here we report that a novel zebrafish mutantcas003 with nonsense mutation in topbp1 gene encoding topoisomerase II β binding protein 1 ( TopBP1 ) exhibits severe definitive hematopoiesis failure . Homozygous topbp1cas003 mutants manifest reduced number of HSPCs during definitive hematopoietic cell expansion , without affecting the formation and migration of HSPCs . Moreover , HSPCs in the caudal hematopoietic tissue ( an equivalent of the fetal liver in mammals ) in topbp1cas003 mutant embryos are more sensitive to hydroxyurea ( HU ) treatment . Mechanistically , subcellular mislocalization of TopBP1cas003 protein results in ATR/Chk1 activation failure and DNA damage accumulation in HSPCs , and eventually induces the p53-dependent apoptosis of HSPCs . Collectively , this study demonstrates a novel and vital role of TopBP1 in the maintenance of HSPCs genome integrity and survival during hematopoietic progenitor expansion .
Hematopoietic stem/progenitor cells ( HSPCs ) possess the capabilities of self-renewal and differentiation into all lineages of mature blood cells [1] . Dysregulated self-renewal of HSPCs is tightly associated with the human blood diseases including leukemia and bone marrow failure ( BMF ) syndrome [2–4] . Previous studies have illustrated that the genes causative for adult hematopoietic diseases virtually play critical roles in the early hematopoiesis [5 , 6] . Therefore , exploring the unknown genetic regulators of HSPCs in the hematopoiesis would give us better understanding of the sophisticated mechanisms of hematopoietic diseases in adults . Recently , zebrafish has emerged as an excellent animal model to study the development of hematopoiesis [7–9] . With multiple unique advantages including external fertilization and development , optically transparent embryos , small size and high fecundity , zebrafish is extraordinarily suitable for the unbiased large scale forward genetics screening to identify novel genes regulating HSPCs self-renewal in the embryonic development [10] . More importantly , the hematopoietic anatomy and the critical transcriptional factors involved in the development of hematopoiesis are highly conserved between zebrafish and mammals [1 , 11] . Similar to mammals , zebrafish hematopoiesis consists of two waves of hematopoiesis , i . e . primitive hematopoiesis and definitive hematopoiesis . The primitive hematopoiesis takes place in the anterior lateral plate mesoderm ( ALPM ) and intermediate cell mass ( ICM ) at ~12–14 somites stage , producing primitive macrophages and erythrocytes , respectively [12] . In zebrafish definitive hematopoiesis , HSPCs originate in the ventral wall of dorsal aorta ( an equivalent of the aorta-gonad-mesonephros [AGM] in mammals ) through endothelium to hematopoietic transition ( EHT ) from 26 hours post fertilization ( hpf ) [13 , 14] , and then colonize in caudal hematopoietic tissue ( CHT , an equivalent to the fetal liver [FL] in mammal ) ( at 2 days post fertilization [dpf] ) , thymus ( at 3dpf ) and ultimately kidney marrow to support adult hematopoiesis ( equivalent to bone marrow ( BM ) in mammal ) ( after 5dpf ) [15 , 16] . During fetal hematopoiesis in CHT , the nascent HSPCs undergo extensive proliferation for the pool expansion to support the embryo development [15] . It has been reported that 95–100% of HSPCs are actively cycling in the mouse fetal liver , whereas most of adult HSPCs are in a quiescent state [17] . During DNA replication , the slowed or stalled DNA replication fork , which is termed as DNA replication stress , occurs frequently due to intracellular and extracellular sources including the by-products of cellular metabolism ( e . g . dNTP misincorporation , reactive oxygen species etc . ) , ultraviolet light and chemical mutagens [18 , 19] . Because the stalled replication forks are vulnerable and the collapse of the forks can result in DNA double strand breaks ( DSBs ) that are deleterious for the genome stability and cell survival , the DNA replication stress-induced DNA damage needs to be efficiently resolved by DNA damage response ( DDR ) pathways [18] . The phosphoinositide kinase-related kinase ataxia telangiectasia mutated ( ATM ) and ATM and Rad3-related ( ATR ) are two important kinases involved in DDR . ATM mainly participates in the DSBs response , whereas ATR is activated by the single-stranded DNA ( ssDNA ) damage and DNA replication stress [20] . Recent studies have shed the light on the association between hematopoietic homeostasis and DDR . DDR impairment can lead to progressive BMF and hematopoietic malignancies [21–23] . Fanconi anemia ( FA ) pathway , which consists of 15 FA genes , mainly participates in repairing the DNA interstrand crosslinks ( ICL ) . Most of the FA genes are associated with the replication fork protection and ATR activation pathway [24 , 25] , and they are causally mutated in BMF or acute myelogenous leukemia [26] . Topoisomerase II β binding protein 1 ( TopBP1 ) is a structurally and functionally conserved protein from yeast to human , which is essential as a scaffold protein in DNA replication initiation and DNA damage checkpoint activation [27–30] . TopBP1 plays a vital role in the DDR , it mainly protects against the ssDNA damage and DNA replication stress through the TopBP1-ATR-Chk1 axis [31–33] . In this process , the stalled replication forks will generate a typical double-stranded DNA-single-stranded DNA ( dsDNA-ssDNA ) structure . Following the replication protein A ( RPA ) coating , TopBP1-associated proteins including Rad9-Rad1-Hus1 ( 9-1-1 complex ) , ATR interaction protein ( ATRIP ) and ATR are recruited to the damage locus , then TopBP1 largely activates the ATR kinase activity through its ATR activation domain ( AAD ) , which triggers the phosphorylation of Chk1 and stabilization of replication forks until the stress is resolved [34–38] . Other TopBP1 interacting components also facilitate the establishment of the TopBP1-ATR-Chk1 axis , including the mediator of DNA-damage checkpoint 1 ( MDC1 ) and BRCA1 interacting protein C-terminal helicase ( BRIP1 , aka , FANCJ ) [39–42] . Although the cellular function of TopBP1 has been established , its physiological role , especially the tissue specific requirement , is still largely unknown . TopBP1 null mice are embryonic lethal due to accumulated DNA damage and reduced cell proliferation , which is phenocopied by TopBP1 W1147R knock-in mice with abrogated AAD domain of TopBP1 [43 , 44] . Moreover , neuronal specific deletion of TopBP1 in mice demonstrates that TopBP1 is essential for neural progenitor cells to survive from the DNA replication stress [45] . Specific disruption of TopBP1 in the lymphoid cells blocks lymphocyte development due to aberrant V ( D ) J rearrangement [46] . However , whether TopBP1 participates in the HSPCs development is still unknown . Here we report a novel zebrafish mutantcas003 , in which HSPCs can be generated normally , but fail thereafter in definitive hematopoiesis . Positional cloning and functional validation indicated that a nonsense mutation-caused C-terminal truncation of TopBP1 was responsible for its subcellular mislocalization and hematopoietic deficits . Disrupted TopBP1-ATR-Chk1 pathway and the accumulation of DNA damage were associated with the HSPCs defect and triggered apoptosis via a p53-dependent pathway . Our findings demonstrate that topbp1 is essential for the HSPCs survival under extensive DNA replication stress during the highly proliferative fetal definitive hematopoiesis .
To explore new genes and regulatory mechanisms in vertebrate definitive hematopoiesis , we carried out a large-scale forward genetics screen on ENU-mutagenized F2 families in zebrafish by whole mount in situ hybridization ( WISH ) using c-myb probe ( a key transcription factor and marker of HSPCs ) [15 , 47] . In 5dpf wild-type zebrafish embryos , c-myb was expressed in all hematopoietic tissues including caudal hematopoietic tissue ( CHT ) , thymus , and kidney ( Fig 1 ) ; whereas homozygous mutantscas003 displayed normal morphogenesis ( Fig 1A–A’ ) , but dramatically decreased c-myb expression in CHT , kidney and thymus ( Fig 1B–B’ ) , suggesting the expansion of HSPCs was defective . To confirm the defective definitive hematopoiesis in mutantscas003 , we further examined the expression of downstream hematopoietic lineage cell markers including ae1-globin ( erythrocyte marker ) , mpx ( granulocyte marker ) , lyz ( macrophage marker ) and rag1 ( lymphocyte marker ) . The expression of all these markers was substantially decreased in the homozygous mutantcas003 embryos at 5dpf ( Fig 1C–G’ ) , which suggested hematopoiesis failure . Recent studies have demonstrated that vasculogenesis and blood flow are essential for HSPCs initiation and maintenance [48 , 49] . We examined the expression pattern of a pan-endothelial cell marker flk1 at 36hpf and an artery vessel marker ephrinB2 at 26hpf respectively , our results revealed that both of them were intact in mutantcas003 ( S1A–S1D Fig ) . Consistently , heart beating rate and blood circulation were comparable between mutantcas003 and sibling control ( S1 and S2 movie ) . In addition , live observation on mutantcas003 , within Tg ( fli1: EGFP ) transgenic background [50] , indicated that the vascular plexus in the CHT region was normal from 2dpf to 5dpf ( S1E–S1L Fig ) . We further investigated the primitive hematopoiesis in mutantcas003 . The WISH analysis data demonstrated that the expression of primitive hematopoietic cell markers were identical between siblings and mutantcas003 at 22hpf , including scl ( hematopoietic progenitor marker ) , gata1 ( erythrocyte progenitor marker ) , pu . 1 ( myeloid progenitor marker ) , lyz , l-plastin ( myeloid cell marker ) and mpx ( S2A–S2L Fig , quantified in M ) . Taken together , we concluded that mutantcas003 displayed specific deficiency in definitive hematopoiesis during zebrafish circulation system development . HSPCs are generated from the ventral wall of dorsal aorta through the endothelia to hematopoietic transition ( EHT ) from 26hpf [13 , 14] , and then migrate to the CHT , a proliferative hematopoietic microenvironment , for pool expansion at 2dpf [15 , 16] . To figure out when the HSPCs defect initiated in mutantcas003 , we performed a time course analysis of c-myb expression from 36hpf to 5dpf . The WISH results demonstrated that the generation of HSC was intact in mutantcas003 as both c-myb and runx1 [51] expression were undisturbed at 36hpf ( Fig 2A–B’ and 2F–G’ ) , and the c-myb expression was still intact in the CHT at 2dpf in mutantcas003 ( Fig 2C–C’ and 2H–H’ ) . However , mutantcas003 displayed reduced c-myb expression in the CHT at 3dpf ( Fig 2D–D’ and 2I–I’ ) , and such defect was more profound at 4dpf ( Fig 2E–E’ and 2J–J’ ) , indicating that the HSPCs proliferation or maintenance was impaired in the CHT of mutantcas003 . To consolidate this discovery , we carried out quantitative RT-PCR analysis on the c-myb mRNA level in zebrafish tails region including CHT from 2dpf to 5dpf . As expected , the c-myb expression level was attenuated from 3dpf to 5dpf ( Fig 2K ) , which was consistent with the results of WISH analysis . To further confirm these findings , we crossed mutantcas003 with Tg ( c-myb: EGFP ) , in which HSPCs could be visualized by EGFP [52] . Statistically significant reduction of EGPF+ cells was observed at 4dpf ( Fig 2L , 2N and 2Q ) and was more severe at 5dpf in mutantcas003 ( Fig 2L , 2O and 2R ) ( Due to the long half-life of EGFP protein , the dynamics of c-myb expression indicated via Tg ( c-myb: EGFP ) was delayed , compared to WISH analysis via c-myb probe [5] ) . Collectively , our data revealed that , in mutantcas003 , neither HSPCs specification in AGM nor their migration to CHT was affected , but their transitory expansion in the CHT was compromised . In order to elucidate the mechanism of hematopoietic failure in mutantcas003 , we carried out positional cloning of the mutant . The mutation was first mapped to chromosome 24 by bulk segregation analysis ( BSA ) . With a high resolution mapping approach , the mutation was revealed to be flanked by two closely linked SSLP markers , L0310_5 and R0310_4 . The flanked region contained four candidate genes: topbp1 ( topoisomerase II β binding protein 1 ) , tmem108 , cdv3 and vps41 ( Fig 3A ) . After sequencing cDNA of all 4 genes , we identified a C to T nonsense mutation in topbp1 gene in mutantcas003 ( Fig 3B ) , and confirmed this result through genomic sequencing . This mutation caused an earlier stop codon before the eighth BRCT ( BRCA1 C-terminus ) domain and a putative C-terminus nuclear localization signal ( NLS ) of TopBP1 protein ( Fig 3C ) . This truncated form of endogenous TopBP1 ( TopBP1cas003 ) protein was further confirmed by immunoblotting analysis of the CHT of heterozygote ( Het cas003 ) and mutantcas003 embryos at 3dpf ( Fig 3D ) . In order to examine whether the disruption of topbp1 was causative for phenotype of mutantcas003 , we injected a validated topbp1 ATG morpholino oligo ( MO ) ( S3A–S3B Fig ) into one-cell stage wild-type embryos to block the translation of endogenous topbp1 mRNA ( Fig 3E ) . Since topbp1 MO acted in a dose-dependent manner ( S3C Fig ) , we applied morpholino microinjection causing no morphologic phenotype in the following studies . Topbp1 morphants manifested severe defective definitive hematopoiesis as that in mutantscas003 from 36hpf to 5dpf , while the primitive hematopoiesis at 22hpf , HSPC generation in AGM at 36hpf and vascular system in CHT at 3dpf were all intact in the morphants ( Fig 3F–G’ and S4A–S4R Fig ) . To further consolidate our findings , we performed rescue assay by ectopic expression of wild-type topbp1 in mutantcas003 . Consistent with previous report on the instability of topbp1 mRNA [53] , ectopic expression of TopBP1 was barely detected at 3dpf after injection of in vitro synthesized topbp1 mRNA into 1-cell stage embryos . In order to overcome this obstacle , we employed a Tol2 transposase-mediated transgenic rescue approach [54] . The ubiquitin promoter ( driving ubiquitous expression ) and the coding sequence of topbp1WT or topbp1cas003 followed by P2A peptide-mCherry fusion protein ( P2A peptide allows self-cleavage of transgenesis efficacy indicator-mCherry without affecting TopBP1 protein ) were constructed into the plasmid containing Tol2 arms ( hereinafter referred to as ubi: topbp1WT and ubi: topbp1cas003 , Fig 3H ) [55 , 56] . After co-injection with Tol2 transposase mRNA and ubi: topbp1WT or ubi: topbp1cas003 constructs into one cell stage mutantcas003 embryos , ubi: topbp1WT driven ectopic expression of wild-type topbp1 could rescue mutantcas003 phenotype at 5dpf ( Fig 3I–3K ) , but not the ubi: topbp1cas003 construct ( Fig 3I–3J and 3L ) . Taken together , the MO phenocopy assays and the wild-type topbp1 rescue assays revealed that the nonsense mutation in topbp1 was the causative mutation in mutantcas003 . Meanwhile , we changed the name of mutantcas003 into topbp1cas003 . To explore how topbp1 affected maintenance of HSPCs in CHT region , we first investigated the expression pattern of topbp1 during embryonic development . WISH analysis data indicated that topbp1 was a maternal mRNA , and was ubiquitously expressed during embryogenesis ( S5A–S5J Fig ) . Previous reports had showed that topbp1 knock-out or knock-down could result in either cell proliferation blockage or cell apoptosis activation [44 , 45] . To investigate the cause of HSPCs abrogation , we conducted cell biology assessment of HSPCs in topbp1cas003 mutants in Tg ( c-myb: EGFP ) transgenic background . Double staining of c-myb and phospho-histone 3 ( pH3 ) showed no significant difference in topbp1cas003 mutants , compared with siblings at 3 . 5dpf ( Fig 4A–D’ , quantified in Q ) , suggesting that the cell cycle of HSPCs was not affected in topbp1cas003 mutants . Furthermore , we performed 5-bromo-2-deoxyuridine ( BrdU ) incorporation assay on HSPCs , BrdU and EGFP double immunostaining results indicated that there was no significant difference in the percentage of BrdU+ HSPCs between siblings and topbp1cas003 mutants at 3 . 5dpf ( Fig 4E–H’ , quantified in R ) . However , TUNEL assay showed a significant increase of apoptotic EGFP+ HSPCs in CHT region of topbp1cas003 mutants , compared with that in wild-type siblings at 3 . 5dpf ( Fig 4I–L’ , quantified in S ) . At 4dpf , the percentage of apoptotic EGFP+ HSPCs was even more significantly increased in topbp1cas003 mutants in comparison with siblings ( Fig 4M–P’ , quantified in S ) , while the number of EGFP+ HSPCs were dramatically decreased ( Fig 2N and 2Q ) . Notably , we could also detect the increased apoptosis in the cranial region and the neural tube in the topbp1cas003 mutants at 3 . 5dpf and 4dpf . Collectively , we concluded that the increased apoptosis in HSPCs was linked to the defective hematopoiesis in topbp1cas003 mutants . To determine how TopBP1 deficiency triggered apoptosis , we firstly checked the expression of several apoptosis-related genes in the CHT regions of topbp1cas003 mutants at 3dpf . The quantitative PCR results showed that the expression of p53 , p21 , cyclin G1 and mdm2 were upregulated in the CHT region of topbp1cas003 mutants , indicating the p53 signaling pathway was activated ( Fig 5A ) . Furthermore , we employed ectopic expression of Bcl2 into topbp1cas003 mutants , which was known to inhibit p53 dependent apoptosis pathway [57] . WISH analysis on c-myb expression showed that bcl2 mRNA could partially restore the c-myb expression in the CHT regions of topbp1cas003 mutants ( 25 out of 43 embryos were partially rescued , S6B–S6D’ Fig , quantified in S6A Fig ) . To confirm the apoptosis in topbp1cas003 HSPCs mainly depended on the p53 pathway , we crossed topbp1cas003 mutant with the tp53M214K mutant ( abbreviated as p53-/- below ) , which had been reported to abrogate p53 function in apoptosis [58] . Further investigation showed that the expression of c-myb in topbp1cas003 mutants was partially rescued in p53+/- heterozygous background at 4dpf ( 3/12 embryos were well rescued , 3/12 embryos were partially rescued , Fig 5B–F’ ) , and the rescue effect was more obviously in p53-/- background at 4dpf ( 7/13 embryos were well rescued , 4/13 embryos were partially rescued , Fig 5B–F’ ) . Taken together , we concluded that the apoptosis of HSPCs in topbp1cas003 mutants was p53-dependent . To further understand the molecular mechanism of HSPC apoptosis which was induced by this particular defective TopBP1 without its 8th BRCT domain and the putative NLS domain , we analyzed the subcellular localization of TopBP1cas003 . Confocal imaging showed that flag-tagged TopBP1WT was predominantly localized in the nucleus of transfected HeLa cells ( Fig 6A , left column ) . However , TopBP1cas003 was mistakenly localized in cytoplasm ( Fig 6A , middle column ) , which was consistent with our previous sequence analysis on the lack of putative NLS in TopBP1cas003 ( Fig 3C ) and immunoblotting analysis on TopBP1WT/TopBP1cas003 protein in cytoplasmic and nucleus fractions of pooled embryos from heterozygote incrossing ( S5L Fig ) . Moreover , addition of SV40 NLS at C terminus of TopBP1cas003 was sufficient to correct TopBP1cas003 subcellular localization defect ( Fig 6A , right column ) . To test whether the hematopoietic deficiency in topbp1cas003 mutants could also be rescued by the nuclear localized TopBP1cas003 , we carried out transient transgenesis of topbp1cas003-NLS or topbp1WT ( as the positive control ) in the topbp1cas003 mutants . WISH results of c-myb at 4dpf indicated that ectopic expression of topbp1cas003-NLS could rescue c-myb expression in topbp1cas003 mutants , as efficient as transgenesis with topbp1WT ( Fig 6B–E’ , quantified in F ) . Collectively , we concluded that the loss of NLS in TopBP1cas003 and the failure of nuclear localization directly caused HSPCs deficiency in topbp1cas003 mutants . Previous studies have demonstrated that TopBP1 plays conserved roles as a scaffold protein that is important for DNA replication and DNA damage response ( DDR ) [27 , 29 , 37] . Since the proliferation of HSPCs was not disrupted in topbp1cas003 mutants ( Fig 4A–H’ ) , it seemed that the function of TopBP1 in DDR instead of DNA replication was responsible for the HSPCs defect in the mutants . Firstly , we checked the activation of TopBP1-ATR-Chk1 pathway in topbp1cas003 mutants and siblings under the hydroxyurea ( HU ) treatment , which was extensively applied to mimic DNA replication stress and could activate ATR-Chk1 axis in mammalian cells and zebrafish embryos [30 , 59 , 60] . The phospho-Chk1 ( pChk1 ) level in CHT region was significantly increased after 250mM HU treatment from 60hpf to 76hpf ( Fig 6G , lane1 and 2 ) . However , the activation of pChk1 was abrogated in topbp1 morphants ( Fig 6G , lane 3 ) . Consistently , we also observed dramatic ablation of pChk1 elevation in the CHT of topbp1cas003 mutants compared with wild-type siblings ( Fig 6H ) . Furthermore , we analyzed protein-protein interaction sites in TopBP1 on the basis of previous biochemical and structural investigations [31 , 41–43 , 61 , 62] . The R122 , R669 and W1156 sites in TopBP1 are involved in Rad9 or MDC1 interaction and ATR activation , respectively . All these sites are highly conserved among zebrafish , human and mouse ( S7 Fig ) , and they are critical for TopBP1-ATR pathway [31 , 41 , 61 , 63] . Transient transgenesis of TopBP1ΔAAD , TopBP1W1156R , TopBP1R122E , TopBP1R669E and TopBP1WT ( as positive control ) in topbp1cas003 mutants was analyzed for hematopoiesis monitored by c-myb WISH . None of these mutated TopBP1 could rescue the hematopoietic failure in topbp1cas003 mutants , compared with TopBP1WT ( Fig 6I ) , indicating that ATR activation function of TopBP1 was essential for HSPCs survival in topbp1cas003 mutants . Taken together , these data implied that the blockage of TopBP1-ATR-Chk1 pathway was correlated to the defective HSPCs in topbp1cas003 mutants . Since TopBP1-ATR-Chk1 axis was disrupted in topbp1cas003 mutants , the unresolved DNA replication stress would result in collapse of replication forks , which could introduce DNA double-stranded breaks ultimately [18] . To check whether the apoptosis of HSPCs was due to the accumulation of DNA damage in CHT region , we carried out fluorescent c-myb WISH analysis and immunostaining with phosphorylated histone H2AX ( γH2AX ) antibody , which was a typical DNA damage marker [64] , from 39hpf to 3 . 5dpf . Interestingly , we couldn’t detect any γH2AX+ cells in AGM region at 39hpf in both topbp1cas003 mutants and siblings , but γH2AX+ HSPCs emerged in CHT region in topbp1cas003 mutants at the same stage ( S8A–S8B Fig ) . Moreover , γH2AX+ HSPCs were accumulated in CHT region of topbp1cas003 mutants afterward ( S8C Fig ) , and they were obviously increased at 3 . 5dpf , ( Fig 7A–H’ , S8C Fig ) indicating the DNA damage was indeed accumulated in HSPCs in topbp1cas003 mutant . In addition , we could also observed several γH2AX+ cells in neuronal tissue ( Fig 7G ) , which was consistent with previous investigation [45] . Furthermore , the immunoblotting of γH2AX within CHT regions of topbp1cas003 mutants at 3dpf also showed an increase of DNA damage ( Fig 7I ) . Collectively , we found that DNA damage was accumulated in HSPCs in topbp1cas003 mutants . To further examine whether the hematopoietic failure was due to the defective DDR upon DNA replication stress in topbp1cas003 mutants , we challenged the embryos with HU . Indeed , high concentration treatment from 52hpf to 76hpf directly caused embryonic lethality in topbp1cas003 mutants ( over 65% ) , however the effect on wild-type siblings was much milder ( <15% ) ( S9A and S9B Fig ) [60] . Furthermore , we carried out a recovery assay in the HU-treated zebrafish embryos ( Fig 7J ) . Interestingly , despite of a suppression by HU treatment , the c-myb expression was recovered in wild-type sibling embryos after challenge removal ( Fig 7K , 7L–L’ , 7N–N’ and 7P–P’ ) . In contrast , the c-myb expression level was not recovered , but decreased further in the HU-treated topbp1cas003 mutant embryos ( Fig 7K , 7M–M’ , 7O–O’ and 7Q–Q’ ) . Taken together , all these observations suggested that the HSPCs in CHT of topbp1cas003 mutants were defective in replicative DNA damage response and they eventually underwent apoptosis through a p53-dependent signaling pathway .
In this study , we reported a novel zebrafish mutant topbp1cas003 , which manifested severe defect in definitive hematopoiesis . The reduction of HSPCs started from 3dpf , which was mainly due to the increased p53-dependent apoptosis , rather than proliferation deficiency . Genetic assessment revealed that a nonsense mutation in topbp1 gene was causative for the hematopoiesis failure . Further investigation revealed that the mutated TopBP1cas003 protein was decreased and mislocalized from nucleus to cytoplasm which compromised the DNA damage response . As a result , it led to accumulated DNA damage that triggered sequential apoptosis of HSPCs in topbp1cas003 mutants . In zebrafish definitive hematopoiesis , HSPCs undergo extensive proliferation in the CHT region around 3dpf , during which the replication stress , characterized by the stalled replication forks , can be induced by various endogenous and exogenous factors [18] . The stalled replication forks will generate typical dsDNA-ssDNA structure , followed with proper loading of RPA , ATR-ATRIP and 9-1-1 complex [37] . Sequential recruitment of TopBP1 can largely activate ATR kinase activity , and the latter will phosphorylate downstream molecules including Chk1 . Activated Chk1 stabilizes the replication forks and arrests cell cycle in order to leave enough time for DNA damage repair machinery to work and to restart the replication fork , so that HSPCs can survive the stress and finish their pool expansion ( Fig 8 ) . The quantitative analysis and WISH results demonstrated that nonsense mutation in topbp1 might lead to nonsense mediated mRNA decay . The expression level of topbp1 was decreased over 80% in the whole embryo and about 50% in CHT of topbp1cas003 mutants ( S5M–S5Q Fig ) . Although around 50% TopBP1cas003 protein remains in CHT , it was mistakenly localized in cytosol , while TopBP1WT was mainly in nucleus to play its role in DDR ( S5L Fig ) . Our results suggest that TopBP1cas003 is decreased and fails in its nucleus entry due to the loss of its C-terminal NLS , abrogating the later ATR/Chk1 activation . In topbp1cas003 HSPCs , the unresolved stalled replication forks would collapse and generate multiple DNA fragile sites , which can induce dsDNA break [18] . As a result , p53-dependent apoptosis is elevated in topbp1cas003 HSPCs , impairing the HSPCs pool severely ( Fig 8 ) . Recently an improved clustered regularly interspaced short palindromic repeats ( CRISPR ) / CRISPR-associated proteins ( Cas9 ) system with custom guide RNAs ( gRNAs ) and a zebrafish codon-optimized Cas9 protein showed high mutagenesis rate in zebrafish , which could even generate biallelic mutations in the F0 generation [65 , 66] . In order to confirm that the deficiency of TopBP1 could disrupt the development of HSPCs , we adapted this optimized CRISPR/Cas9 system to obtain other topbp1 zebrafish mutants ( S10A–S10B Fig ) . Some of the topbp1 Cas9 injected wild-type embryos displayed dramatically decreased c-myb expression as same as topbp1cas003 mutant at 4dpf ( S10C–S10D’ Fig ) . And this phenotype could be reached in higher efficiency when the injected embryos were generated from the outcross between topbp1cas003 heterozygote and wild-type fish ( S10E–S10F’ Fig ) . Conclusively , these data provided additional evidence that definitive HSPCs were defective in the TopBP1 loss-of-function embryos . It is an intriguing finding that topbp1 plays an essential role in proliferative tissues , especially in the definitive hematopoiesis without affecting the morphogenesis at the early stage , whereas its transcripts were ubiquitously distributed in the embryogenesis ( S5 Fig ) , and TopBP1 knockout mice were reported to be lethal at the peri-implantation stage [44] . The WISH analysis showed maternal expression of topbp1 ( S5A Fig ) , suggesting that homozygote topbp1 mutant embryos can inherit wild-type topbp1 mRNA from the female parents to support its early development until zygotic topbp1 expresses latter in the development . Nevertheless , we attempted to figure out whether topbp1 was expressed and functional in the HSPCs . Quantitative PCR analysis on the CD41+ cell population in the tail region of Tg ( CD41: EGFP ) embryos , which was reported to be an enriched population of HSPCs at 5dpf [67 , 68] , showed that the level of topbp1 mRNA was 3-fold enriched in CD41+ cells , compared to cells in the whole tails , demonstrating its expression in HSPCs ( S5K Fig ) [5] . Furthermore , due to the lack of definitive hematopoiesis-specific promoter , we used hemangiogenic promoter lmo2 , which was also expressed in definitive HSPCs , to drive the ectopic expression of wild-type topbp1 into topbp1cas003 mutants [52 , 69] , we could indeed observe the expression of mCherry driven by lmo2 promoter in CHT region at 5dpf , and this construct could partially rescue the HSPCs deficits at 5dpf ( S11 Fig ) . In addition , the vascular plexus in CHT region was normal in topbp1cas003 mutants or morphants from 2dpf to 5dpf ( S1E–S1L Fig and S4O–S4R Fig ) , and low dose microinjection of topbp1 morpholino was sufficient to induce definitive hematopoiesis deficits in CHT without affecting the primitive hematopoiesis and vascular system in wild-type embryos ( S3C–S3D Fig , S4 Fig ) . Taken all these data together , we concluded that TopBP1 played an essential and HSPC-intrinsic mechanism during definitive hematopoiesis . It is intriguing that whether the truncated TopBP1 can potentially function as a dominant negative protein . Ectopic expression of cas003 mutant form of TopBP1 ( TopBP1cas003 ) driven by ubiquitin promoter was performed in wild-type fish , and it did not cause defective definitive hematopoiesis ( S12 Fig ) . The possible reason for this phenomenon was that the mutated TopBP1 could not enter nucleus to compete with wild-type TopBP1 . Meanwhile , the hematopoietic phenotype of topbp1cas003 heterozygotes was checked , and no HSPCs defect was observed . Taking these results together , we concluded that TopBP1cas003 could not function as a dominant negative form . In definitive hematopoiesis , nascent HSPCs seldom proliferate in AGM region , but they become active in cell cycle and undergo extensive proliferation in CHT region supported by niche cells , meanwhile , they have to overcome DNA replicative stress [13 , 15 , 18] . BrdU incorporation assays within Tg ( c-myb: EGFP ) embryos confirmed that HSPCs underwent high proliferation at a constant rate from 2dpf to 5dpf , although the expansion of neural tube cells was gradually attenuated ( S13 Fig ) . As a result , the defect in HSPCs was more profound after 3dpf in the topbp1cas003 mutant . Consistently , we indeed found obvious accumulation of γH2AX positive cells ( 2 . 5dpf ) and increased apoptotic cells ( 3 . 5dpf ) in cranial and neuron tube tissues of topbp1cas003 mutant , which was in agreement with previous observations in neuron-specific TopBP1 knock-out mice [45] . Besides , some of homozygote topbp1 mutant embryos developed smaller head and eyes after 6dpf , and all of them eventually died around 10–20 dpf . Previous works within zebrafish mutants revealed several genes and pathways which were critical for the HSPCs development in CHT region , including genes involved in mitotic spindle assembly , maintenance of centrosome integrity and mitotic progression; pre-mRNA processing; sumoylation of genes participating in DNA replication or cell cycle regulation [5 , 70 , 71] . All these genes were indispensable for cell to complete proliferation or division . Because the HSPCs were highly proliferative in CHT , these data depict a picture that the HSPCs in fetal stage are extremely sensitive to the disruption of genes participating in various processing to complete cell division successfully and faithfully . As the DDR pathway is essential for genomic fidelity and stability during DNA replication , our work revealed that DDR pathway is also critical for HSPCs development in fetal stage . It has been reported that Fanconi anemia pathway is critical for the repair of DNA cross-link damage [26] . Biallelic mutations in any of 15 FANC genes will result in Fanconi anemia ( FA ) , which can most frequently develop into inherited bone marrow failure ( BMF ) syndrome [72] . The work of Raphael Ceccaldi et al . revealed that the FA patients showed profound HSPCs defect before the onset of BMF [73] . The p53-p21 axis , triggered by replicative stress , was highly elevated in FA HSPCs , and the p53 silence can rescue hematopoietic deficits [73] . They also pointed out that p53 activation , caused by unresolved cellular abnormality , may be the signaling mechanism for inherited BMF , and the p53 activation was commonly found in other types of inherited BMF syndromes , such as Diamond Blackfan anemia ( DBA ) and dyskeratosis congenital ( DC ) [73] . HSPCs in topbp1cas003 mutants manifested similar features as that in FA ( Fig 5 ) , whether topbp1 could be a putative pathogenic gene in human BMF syndrome needs further investigation . Zebrafish fancd2 morphant exhibited developmental abnormalities and p53-dependent apoptosis , however its hematopoietic phenotype had not been extensively investigated [57] . The emi1 homozygous mutants showed disrupted genomic integrity and hematopoiesis failure [74] . Studies on topbp1cas003 mutants revealed that DNA damage and apoptosis signaling was accumulated in the HSPCs of topbp1cas003 homozygous embryos , which linked to the hyper-activated p53-p21 axis ( Fig 5 ) and failed ATR/Chk1 activation ( Fig 6 ) . Furthermore , TopBP1-involved c-myb regulated DDR pathway was proposed by recent studies on castration-resistant prostate cancer [75] . HU treatment of the developing zebrafish further emphasized the importance of DNA damage response and repair pathway for HSPCs survival during high proliferation stage . Collectively , we demonstrated a novel and essential role of TopBP1 in HSPCs during their rapid proliferation in fetal hematopoiesis . Due to the dramatic definitive hematopoiesis phenotype in embryogenesis , topbp1cas003 mutants provide a unique model for the mechanism study and small molecular chemical screen on BMF-like hematopoiesis failure , which is caused by defective replicative DNA damage response .
The zebrafish facility and study were approved by the Institutional Review Board of the Institute of Health Sciences , Shanghai Institutes of Biological Sciences , Chinese Academy of Sciences ( Shanghai , China ) , and zebrafish were maintained according to the guidelines of the Institutional Animal Care and Use Committee . Wild-type ( WT ) zebrafish strains Tubingen ( TU ) and WIK , the transgenic zebrafish line Tg ( c-myb: EGFP ) [52] , Tg ( fli1: EGFP ) [50] , Tg ( CD41: EGFP ) [76] , the mutant zebrafish line tp53M214K/M214K [58] were maintained as previously described [77] . For the forward genetics screen , WT TU zebrafish line was treated with ethylnitrosourea ( ENU , Sigma ) to generate mutants , the screen approach was performed as previously described [78 , 79] . The desired mutants within F3 generation were identified by the whole-mount in situ hybridization ( WISH ) using c-myb probe at 5dpf . For the chemical treatment , the hydroxyurea ( HU , Sigma ) was dissolved with distilled water into 1M and stored at -20℃ . The embryos were treated with 250mM HU as the indicated procedures in the egg water at 28 . 5℃ [59 , 60] . To prevent the formation of melanin pigment , the embryos were incubated in egg water containing 0 . 045% 1-phenyl-2-thiourea ( PTU , Sigma ) after gastrulation stage . The embryos were collected at the desired stages [80] . Positional cloning was carried out with WIK line as previously described [81] . Firstly , the mutation was mapped to chromosome 24 by bulk segregation analysis ( BSA ) with simple sequence length polymorphism ( SSLP ) markers . Through high resolution mapping analysis on 1041 mutants , the mutation was finally flanked by two SSLP markers , L0310_5 and R0310_4 . The candidate genes in this range were sequenced with wild type sibling and mutant cDNA , and the putative mutation was confirmed by genomic DNA sequencing . The primers used in the positional cloning were provided in supplemental S1 Table . Most experiments in this study were carried out with the embryos generated by the incross of mutantcas003 heterozygote pairs ( TU/WIK background ) used in the positional cloning if possible . The mutants can be identified by flanked SSLP markers , such as Z9852 and R0306_4 . Alternatively , the mutants can be distinguished by restriction fragment length polymorphism ( RFLP ) using EcoP15I ( NEB ) , the RFLP primers were provided in supplemental data ( S1 Table ) . To construct Tol2 transgenesis vectors , the ubiquitin promoter [55] or lmo2 promoter [69] followed by P2A [56] and in-frame mCherry was cloned into modified Tol2 backbone [82] . The vectors were referred as pUbi-Tol2 or pLmo2-Tol2 below . The genes of interest can be inserted between the promoter and P2A . Zebrafish topbp1WT or topbp1cas003 were amplified and inserted into pUbi-Tol2 or pLmo2-Tol2 vectors . To generate the mutated forms of topbp1 , the mutagenesis was carried out following QuikChange mutagenesis kit instruction using pUbi-topbp1WT-Tol2 plasmid as the template . The region of TopBP1 ( 984–1206 ) are the putative ATR activation domain ( AAD ) between BRCT6 and BRCT7 . In TopBP1ΔAAD , the coding sequence of TopBP1 ( 1083–1159 ) containing conserved RQLQ and WDDP sequences are deleted [31] . The fragment of topbp1 ( -9–692 ) was amplified and inserted into the pCS2+ vector for in situ probe preparation . To construct topbp1 MO effect evaluation plasmid , a DNA fragment containing topbp1 ATG MO targeting site was inserted into the upstream of EGFP coding region in pCS2+ . Zebrafish topbp1WTand topbp1cas003 were cloned into pCMV4-FLAG-4 vector ( Sigma ) . The SV40 NLS ( nuclear localization signal ) sequence ( 5’-CCAAAAAAGAAGAGAAAGGTA-3’ ) [83] was firstly cloned into pCMV4-FLAG-4 vector in the 3’ end of FLAG tag , and then the topbp1cas003 sequence was inserted into the pCMV4-FLAG-NLS plasmid . All of the primers used were listed in S1 Table . The mRNA was synthesized in vitro by SP6 mMessage mMachine Transcription Kit ( Ambion ) . The topbp1 gRNA was synthesized as described [66] . The information of the topbp1 gRNA target site was shown in S1 Table . The zebrafish optimized Cas9 mRNA was synthesized in vitro from the pCS2-nCas9n plasmid ( addgene , #47929 ) as described [65] . bcl2-egfp mRNA ( ~100pg ) was injected into 1-cell stage embryos [54] . For the ectopic-expression , Tol2 transposon-mediated transient transgenesis was applied and performed as previously described [84] . A series of topbp1 transgene constructs within Tol2 vectors ( ~40 ng/μl ) were mixed with transposase mRNA ( ~60 ng/μl ) and 0 . 2 M KCl , and then injected into 1-cell stage embryos , respectively [85] . The volume of the mixture injected was about 0 . 5nL . The topbp1 ATG morpholino oligo ( MO ) ( 5’-CCTTGCTGGCTTTCGACATGGTGAC-3’ ) and control morpholino ( 5’- CCTCTTACCTCAGTTACAATTTATA-3’ ) were synthesized by Gene Tool company and was injected into 1-cell stage embryos . For Cas9 assay , topbp1 gRNA ( 50pg ) and Cas9 mRNA ( 150pg ) were co-injected into one-cell stage embryos . The T7EI assay was performed as described [65] . c-myb , runx1 , ae1-globin , mpx , lyz , rag1 and topbp1 probes were transcribed in vitro by T3 or T7 polymerase ( Ambion ) with Digoxigenin RNA Labeling Mix ( Roche ) . One color WISH was performed as described previously [54] . Images were photographed by the Nikon SMZ1500 microscope with Nikon DXM 1200F CCD or Olympus SZX16 microscope with Olympus DP80 CCD . c-myb RNA and immuno-fluorescence double staining was carried out as described previously [70] . For the immunostaining , rabbit anti-pH3 antibody ( 1:500 , Santa Cruz ) and rabbit anti-γH2AX antibody ( 1:500 , gift from Dr . James Amatruda , University of Texas Southwestern ) were used . The embryos were stained with goat-Alexa Fluore488-conjugated anti-rabbit secondary antibody ( 1:500 , Invitrogen ) . DAPI ( 1:500 , Beyotime ) staining was carried out along with the secondary antibody incubation if necessary . The 3 . 5dpf topbp1cas003 mutant/Tg ( c-myb:EGFP ) or sibling embryos were soaked in egg water containing 10mM BrdU ( Sigma ) /15% DMSO for 30 minutes at 28 . 5℃ or injected with 1nL 30mM BrdU into the yolk sac . Then they were transferred into fresh egg water and incubated for 2 hours . After fixation in 4% paraformaldehyde ( PFA , Sigma ) , the embryos were dehydrated with methanol and stored at -20℃ overnight . For BrdU immunostaining , the rehydrated embryos were digested with Proteinase K ( 12 μg/ml , Roche ) at 30℃ for 28 minutes and treated with acetone at -20℃ for 30 minutes . After re-fixation with 4% PFA , the embryos were blocked with the block solution ( PBS + 0 . 3% Triton-X -100 +1% DMSO+ 10 mg/ml BSA+10% normal goat serum ) for 2 hours at RT . The embryos were then incubated with anti-GFP Rabbit Serum ( 1:500 , Invitrogen ) followed by goat-Alexa Fluore488-conjugated anti-rabbit secondary antibody ( 1:500 , Invitrogen ) incubation . 2N HCl was used to treat the embryos for 1 hour at room temperature ( RT ) . After that , the embryos were stained with mouse anti-BrdU primary antibody ( 1:50 , Roche ) and goat-Alexa Fluore546-conjugated anti-mouse secondary antibody ( 1:500 , Invitrogen ) . TUNEL assay was performed with the In Situ Cell Death Detection Kit TMR red ( Roche ) . Similar to the BrdU immunostaining , 3 . 5dpf and 4dpf topbp1cas003 mutant/Tg ( c-myb:EGFP ) or sibling embryos were fixed with 4% PFA . After methanol dehydration , rehydration , Proteinase K digestion and acetone treatment , the embryos were permeated with permeabilisation solution ( 0 . 5% Triton X–100 , 0 . 1% sodium citrate in PBS ) at RT for 4 hours . Then the embryos were stained with the TUNEL Kit ( 100ul , enzyme: labeling solution = 1:9 ) at 37℃ for 2 hours . Finally , the EGFP immunostaining was carried out as described above . The CD41+ cells were sorted from the tails of Tg ( CD41: EGFP ) embryos including the CHT region at 5dpf as previously described [86 , 87] . The total RNA was extracted from TRIzol ( invitrogen ) dissolved zebrafish whole embryos or the tails including CHT region or the sorted cells , and then transcribed into cDNA by PrimerScript RT Master Mix ( TaKaRa ) . The quantitative PCR was carried out with SYBR Green Real-time PCR Master Mix ( TOYOBO ) with ABI 7900HT real-time PCR machine , and analyzed with Graphpad 5 . 1 software . The primers used were listed in S1 Table . HeLa and HEK293T cells were maintained in DMEM with 10% Fetal Bovine Serum ( FBS ) and penicillin-streptomycin antibiotics ( 1:100 ) . Plasmid transfection was carried out with Lipofectamine 2000 ( Invitrogen ) according to manufacturer’s instruction . The immunostaining was carried out in HeLa cells as previously described [70] . FLAG-topbp1WT , FLAG-topbp1cas003 and FLAG-topbp1cas003-NLS plasmids were transfected into HeLa cells . Mouse anti-FLAG primary antibody ( 1:1000; Genomics Technology ) and goat-Alexa Fluore488-conjugated anti-mouse secondary antibody ( 1:500 ) were used for immunostaining . DAPI ( 1: 500 , Beyotime ) was applied for nucleus staining . To extract the protein from the cell line , the cells were homogenized directly with 2 X SDS sample buffer and boiled for 5 minutes at 95℃ . To obtain fish protein from the CHT region , the tails of embryos including the CHT region were cut down , then ultrasonicated in RIPA lysis buffer ( 50mM Tris ( pH7 . 4 ) , 150mM NaCl , 1% NP-40 , 0 . 5% sodium deoxycholate , 0 . 1% SDS ) . After centrifugation at 12000rpm for 15 minutes , the supernatant was mixed with 2XSDS sample buffer and boiled for 10 minutes . Cytoplasmic and nuclear extracts were prepared from the 3dpf embryos with Nuclear and Cytoplasmic Protein Extraction Kit ( Beyotime ) according to the manufacturer’s instruction . The immunoblotting was carried out as previously described [85] , with rabbit anti-phospho-Chk1 ( Ser345 ) ( 133D ) antibody ( Cell Signaling Technology ) , rabbit anti-γH2AX antibody , rabbit anti-zebrafish TopBP1 antibody ( generated by 840–940 amino acid of zebrafish TopBP1 protein as antigen ) , mouse anti-GAPDH antibody ( 1D4 ) ( Santa Cruz ) , mouse anti-alpha-tubulin antibody ( Sigma ) or rabbit anti-Histon3 ( H3 ) antibody ( Abcam ) . Images of zebrafish immunofluorescence staining or live transgenic embryos were taken by Olympus FV1000 scanning confocal microscope . The embryos were mounted in 1% low-melt agarose in a self-made 35mm coverslip-bottom dish . The confocal images were captured with an UPLSAPO 20X or 60X objective . To obtain images of HeLa cells immunostaining , the slides were directly immersed in the PBS solution in a 10cm dish . The images were captured with an UPLSAPO 40X objective . The transient transgenesis embryos and embryos for bright field imaging were anesthetized with 0 . 03% Tricaine ( Sigma-Aldrich ) , mounted in 3% methylcellulose and imaged using a Zeiss Axio Zoom . V16 microscope equipped with a Zeiss AxioCam MRm digital camera . Data were analyzed with the Graphpad Prism 5 software using the two-tailed Student’s t-test . The plot error values were calculated by standard error of the mean ( SEM ) . All data in this study were repeated for at least twice . | The rapidly proliferating hematopoietic stem/progenitor cells ( HSPCs ) require well-established DNA damage response/repair pathways to resolve the DNA replication stress-induced DNA damage , which is deleterious for the genome stability and cell survival . Impairment of these pathways could lead to the progressive bone marrow failure ( BMF ) and hematopoietic malignancies . Here we reported a novel function of topoisomerase II β binding protein 1 ( TopBP1 ) in definitive hematopoiesis through characterizing zebrafish mutantcas003 with a nonsense mutation in topbp1 gene encoding TopBP1 . The homozygous topbp1 mutants manifested decreased HSPCs during their pool expansion in the caudal hematopoietic tissue ( CHT , an equivalent of the fetal liver in mammals ) due to the p53-dependent apoptosis . Further investigation revealed that the deficient TopBP1-ATR-Chk1 pathway upon DNA replication stress in topbp1 mutants led to accumulated DNA damage and further affected HSPCs survival . These studies therefore emphasized the importance of topbp1 function as well as DNA damage response pathways during the fetal HSPC rapid proliferation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | TopBP1 Governs Hematopoietic Stem/Progenitor Cells Survival in Zebrafish Definitive Hematopoiesis |
Many components of Wnt/β-catenin signaling pathway also play critical roles in mammary tumor development , yet the role of the tumor suppressor gene APC ( adenomatous polyposis coli ) in breast oncongenesis is unclear . To better understand the role of Apc in mammary tumorigenesis , we introduced conditional Apc mutations specifically into two different mammary epithelial populations using K14-cre and WAP-cre transgenic mice that express Cre-recombinase in mammary progenitor cells and lactating luminal cells , respectively . Only the K14-cre–mediated Apc heterozygosity developed mammary adenocarcinomas demonstrating histological heterogeneity , suggesting the multilineage progenitor cell origin of these tumors . These tumors harbored truncation mutation in a defined region in the remaining wild-type allele of Apc that would retain some down-regulating activity of β-catenin signaling . Activating mutations at codons 12 and 61 of either H-Ras or K-Ras were also found in a subset of these tumors . Expression profiles of acinar-type mammary tumors from K14-cre; ApcCKO/+ mice showed luminal epithelial gene expression pattern , and clustering analysis demonstrated more correlation to MMTV-neu model than to MMTV-Wnt1 . In contrast , neither WAP-cre–induced Apc heterozygous nor homozygous mutations resulted in predisposition to mammary tumorigenesis , although WAP-cre–mediated Apc deficiency resulted in severe squamous metaplasia of mammary glands . Collectively , our results suggest that not only the epithelial origin but also a certain Apc mutations are selected to achieve a specific level of β-catenin signaling optimal for mammary tumor development and explain partially the colon- but not mammary-specific tumor development in patients that carry germline mutations in APC .
Breast cancer is one of the most common malignancies in women in Western countries and it is the cause of death in approximately 20% of all females who die from cancer . Breast epithelium is a dynamic organ capable of rapid proliferation and functional differentiation upon pregnancy and lactation , followed by involution and remodeling at the end of each lactation period . The adult mammary gland consists of secretory alveoli organized into lobules and interconnected by a system of branching ducts . The entire mammary epithelium is enveloped by a basement membrane and embedded in a fatty connective tissue called the mammary fat pad . In the ducts and alveoli , the mammary epithelium is organized into two layers , a basal layer of myoepithelial cells and a luminal epithelial layer . The myoepithelial cells , like other basal epithelial cells , express basal keratins ( in particular , K5 and K14 ) , P-cadherin , and the transcription factor p63 [1] . They also contain smooth muscle-specific proteins , including the α-smooth muscle actin ( α-SMA ) , which confer contractility . By contrast , luminal cells express K8 and K18 , which are characteristics of simple epithelia and when fully differentiated , secrete milk proteins [1] . The molecular mechanisms of the initiation of breast cancer are well studied . Mutations in BRCA1 and BRCA2 result in increased susceptibility to breast cancer [2] and mutations in TP53 are found to be common in late stages of this cancer [3] . It has been shown that dysregulation of the Wnt signaling pathway is an important contributor to the initiation of breast cancer [4] . Adenomatous Polyposis Coli ( APC ) is a member of the Wnt/β-catenin signaling pathway that is involved in the maintenance of the progenitor cell population in the skin , intestine and other tissues . Mutations and/or altered expression in the tumor suppressor gene APC are frequently found in sporadic breast cancers [5]–[7] which implicates its role as a tumor suppressor in mammary epithelium . In mouse , activation of Wnt/β-catenin signaling in the mammary epithelium either by mutation in Apc ( GenBank NM_007462 ) or by stabilization of β-catenin ( NM_007614 ) , contributes to tumorigenesis . For example , mice heterozygous for germline mutation in Apc ( ApcMin ) spontaneously develop mammary tumors , although at a significantly lower incidence than intestinal tumors [8] . Transient expression of an activated form of β-catenin in secretory luminal epithelium driven by the MMTV promoter leads to both mammary gland hyperplasia and mammary adenocarcinoma [9] . Similarly , expression of a transcriptionally active form of β-catenin lacking the N-terminal 89 amino acids ( ΔN89 β-catenin ) results in precocious development , differentiation , and neoplasia in both male and female mouse mammary glands [10] . The K5 promoter-driven expression of stabilized N-terminally truncated β-catenin ( ΔN57 β-catenin ) in the basal epithelial layer of the mammary gland , led to basal-type mammary hyperplasia and invasive carcinomas [11] . Contrasting results have been obtained when Wnt/β-catenin signaling pathway was stably activated constitutively in luminal cells of mammary epithelium using Cre-loxP technology . The stabilization of β-catenin , aided either by Cre-mediated oncogenic activation of β-catenin or Apc deficiency , induced transdifferentiation into epidermis and squamous metaplasia of the mammary epithelium but failed to induce neoplasia [12] , [13] . Apc deficient luminal epithelium developed acanthomas only in the additional absence of Tcf-1 [13] . Together , these results indicate a key role for Apc in both mammary gland development and tumorigenesis , most likely through activation of β-catenin signaling , but it is still unclear why the variation in methods of β-catenin signaling activation can produce different phenotypes in mammary glands . These results suggest that the timing and the cell types in which the Apc mutations occur might be important for breast cancer development . To better understand how Apc inactivation in the mammary epithelium results in cancer , we crossed mice carrying a floxed allele of Apc to K14-cre and whey acid protein ( WAP ) -cre transgenic mice . K14-expression starts embryonically in cells that give rise to both basal and luminal cells of mammary gland , while WAP expression is restricted to adult females following pregnancy and lactation . We show here that K14-mediated Apc heterozygosity directly resulted in mammary adenocarcinoma development , but WAP-mediated Apc deficiency resulted in severe squamous metaplasia and not readily in neoplasia . The expression of both luminal and myoepithelial lineage markers , as well as the presence of the common initiating somatic Apc mutation in histologically distinct regions of a tumor , is in line with the progenitor cell origin of K14-cre; ApcCKO/+ tumors . The remaining wild-type allele of Apc in these tumors harbored truncation mutation in a specific region of the gene , which seems to be selected for mammary tumorigenesis . These results show that the timing and cell type in which the critical mutational events occur and the level of resultant activation of the β-catenin signaling cascade are critical for the initiation of mammary tumor development .
ApcΔ580/+ mice , a germline knockout strain derived from the Apc conditional mice die primarily due to development of multiple intestinal tumors [14] . We have found that these mice can occasionally develop mammary tumors , as in ApcMin/+ mice [8] although at a low incidence ( 14 . 3% , 3 of 21 ) . To further study the role of Apc in mammary tumor development without being hindered by the intestinal tumorigenesis , we induced Apc mutations specifically in mammary epithelium using either K14 or WAP promoters . We have previously shown that homozygous loss of Apc in K14-expressing embryonic cells results in abnormal skin phenotype associated with aberrant development and squamous metaplasia in many epithelial-derived tissues including teeth and thymus , and die prior to weaning [14] . In contrast to K14-cre; ApcCKO/CKO mice , the K14-cre; ApcCKO/+ mice were phenotypically normal at birth , but upon aging showed decreased survival primarily due to mammary tumor susceptibility in female mice ( Table 1 , Figure 1A–G ) . The K14-cre; ApcCKO/+ female mice ( n = 19 ) had a median survival of 15-months . We were able to carefully analyze 17 of these mice for pathology . The differences in survival are statistically significant between the K14-cre; ApcCKO/+ female mice and the cre-negative ApcCKO/+ and ApcCKO/CKO female mice ( p<0 . 02 , log rank test ) . A large proportion of K14-cre; ApcCKO/+ females invariably developed mammary tumors with focal squamous metaplasia ( 13 of 17 , 76 . 5% , Figure 1B–G ) and the mice were sacrificed when their tumors were over 2 cm in diameter . Of the four mammary tumor-free female mice , three died of hepatoma , histiocytic sarcoma and myxosarcoma , respectively and one succumbed to severe dermatitis . We extensively performed a full histological autopsy on seven mammary tumor-bearing mice , of which two had lung metastasis . The mammary tumor susceptibility was also observed in K14-cre; ApcCKO/+ female mice backcrossed to C57BL/6 ( K14-cre; ApcCKO/+-B6 , n = 8 ) . At the time of analysis , half of them ( 4 of 8 , 50% ) developed mammary tumors before reaching 12-months of age while the other half remained tumor-free for over 12-months . Both WAP-cre positive heterozygous and homozygous ApcCKO mice were born in the expected Mendelian ratio with no bias towards either sex . All mice were phenotypically normal at birth , developed normally and were fertile . However , litters from WAP-cre;ApcCKO/CKO mothers could not thrive . When litters were transferred to foster mothers , these litters survived and developed normally , suggesting that it was due to lack of appropriate milk production by WAP-cre;ApcCKO/CKO mothers and consequent inability to nurse their litters properly . This observation is in agreement with the BLG-cre mediated inactivation of Apc [13] . Both WAP-cre positive heterozygous and homozygous ApcCKO female mice have been allowed to pass through four complete lactation cycles with the exception of the two mice that underwent three and were monitored up until they were 18-months of age . Nulliparous females of the same genotypes were also monitored as their controls . Unlike in the K14-cre;ApcCKO/+ nulliparous female mice that developed mammary tumors spontaneously , neither mated WAP-cre positive ApcCKO/CKO nor ApcCKO/+ mice showed mammary tumor susceptibility ( Table 1 ) . There were hardly any differences in either the survival or tumorigenicity between multiparous and nulliparous WAP-cre positive females of either ApcCKO genotypes and all of them lived as long as Cre negative controls ( Table 1 ) . Mammary tumors were occasionally observed in all four groups of WAP-cre positive females ( Table 1 ) . Mammary tumors developed in two out of 14 ( 21 . 4% ) WAP-cre;ApcCKO/CKO and one of 6 ( 16 . 7% ) WAP-cre;ApcCKO/+ multiparous females , whereas those in nulliparous females were one of 10 ( 10% ) and one of eight ( 12 . 5% ) , respectively . Examination of mammary glands from the aged multiparous WAP-Cre;ApcCKO/CKO female mice revealed severe squamous metaplasia in mammary glands that explains the inability of these females to produce milk ( Figure S1A , B ) , whereas those of the age matched multiparous WAP-Cre;ApcCKO/+ , nulliparous WAP-Cre;ApcCKO/+ and nulliparous WAP-Cre;ApcCKO/CKO females had histologically virginal state without any acini development . The extent of metaplasia was so severe in multiparous WAP-Cre;ApcCKO/CKO mice that almost all acini had squamous metaplasia , some with mineralization , showing osteometaplasia ( Figure S1B ) . This observation is analogous to BLG-cre-mediated inactivation of Apc [13] and WAP-cre-mediated activation of oncogenic β-catenin [12] , further supporting that the homozygous mutations of Apc in mammary epithelium perturbs normal mammary differentiation and causes transdifferentiation , but does not readily result in tumorigenesis . These results suggest the timing and perhaps the cell type in which the Apc mutations occur is critical for mammary tumor development . Tumors arising from stem or progenitor cells may show mixed lineage differentiation [15] . K14-expression starts embryonically , and some of those cells give rise to both basal and luminal cells of mammary gland [16] , while WAP expression starts in adult luminal mammary epithelium following pregnancy and lactation [17] . To investigate whether Apc mutation-induced tumors arising from K14-cre positive cells and WAP-cre positive cells have similar histology and lineage differentiation , these tumors were histologically examined and stained for both K8 , a marker for luminal epithelial cells , and α-SMA , K14 and p63 , markers for basal myoepithelial cells . The histology of mammary tumors that developed in the germline knockout strain , ApcΔ580/+ mice were similar to those described for ApcMin/+ mammary tumors [18] . All three of them were pilar tumors with extensive keratinization , and were adjacent to basosquamous components . The mammary tumors developed in K14-cre; ApcCKO/+ mice , either in mixed or C57BL/6 background , exhibited a variety of histological patterns within a tumor similar to those found in other Wnt Pathway tumors [18] . Most of the tumors were adenocarcinomas with focal squamous metaplasia ( Figure 1B–G , 2 ) . Squamous metaplasia may be extensive as in pilar tumors or scattered as multiple foci . The most common histological pattern observed in 16 K14-cre; ApcCKO/+ mammary tumors from mice in the mixed background were acinar ( Figure 1E , 2A–E ) , often associated with basosquamous ( Figure 1F , 2F–J ) and pilar ( Figure 1D , 2K–O ) components but only occasionally with undifferentiated component ( Figure 1G , 2P–T ) . Upon backcrossing to C57BL/6 , the K14-cre; ApcCKO/+ mammary tumors ( n = 4 ) developed were primarily composed of basosquamous and pilar histological types with extensive keratinization , similar to ApcΔ580/+ tumors , and acinar histology was no longer observed . All K14-cre; ApcCKO/+ mammary tumors exhibited multiple histological patterns within a tumor , some more prominent than the others . In two cases , the tumors appeared grossly biphasic with distinct keratinized and solid portions ( Figure 1B , C ) . Histologically , the keratinized portion had pilar structures with a number of keratinizing cysts ( Figure 1D , 2K ) , and the solid cellular portion had the acinar pattern ( Figure 1E , 2A ) when the tumor derived from a mixed background K14-cre; ApcCKO/+ mouse . A similar biphasic growth was also observed in two mammary tumors from K14-cre; ApcCKO/+-B6 mice , but the solid portion had basosquamous pattern . Immunochemical examination of these tumors revealed that these tumors are composed of both luminal and myoepithelial cells , with α-SMA and K14 positive myoepithelial cells forming a single layer around the K8-positive tumor cells in a well-organized structure as in normal ducts ( Figure 2B–D , G–I ) . The immunochemical patterns for K14 and K8 shifts to those of epidermis in pilar structures ( Figure 2L–N ) . Only in undifferentiated mammary tumor components , the expression of both lineage markers was lost ( Figure 2Q–S ) . These tumors were highly proliferative as determined by Ki67 staining ( Figure 2E , J , O , T ) , although the proliferation pattern of the pilar tumors was restricted to basal layer , analogous to that of hair follicles . Strong positivity for Tcf/β-catenin target genes , Myc and cyclin D1 , demonstrating the activation of the Wnt/β-catenin pathway were also observed in these tumors ( Figure S2A–G ) . Like in many other mouse mammary tumor models , K14-cre; ApcCKO/+ mammary tumors were negative for hormone receptors , Estrogen Receptor ( ER ) and Progesterone Receptor ( data not shown ) . Most WAP-cre induced tumors , of which three were from WAP-cre; ApcCKO/CKO and two were from WAP-cre; ApcCKO/+ mice , were histologically acinar or glandular-like ( Figure 2U–Y ) . Interestingly , all except one were only positive for K8 and did not show defined expression of myoepithelial markers , K14 and p63 , as in K14-cre; ApcCKO/+ mammary tumors , and α-SMA expression was very diffuse and aberrant ( Figure 2V–X ) . They also had squamous metaplasia where K14 expression is observed but with less defined structures than those observed in K14-cre; ApcCKO/+ tumors ( Figure S1C , D ) . These observations suggest that K14-cre; ApcCKO/+ mammary tumors derived from either stem or progenitor cells of the mammary gland while WAP-cre induced tumors derived from more differentiated cells of mammary luminal cells . It is known that in Apc-mediated tumorigenesis , an important initial event is the loss or mutation of the second copy of the Apc locus . We examined the status of Apc in 20 K14-cre; ApcCKO/+ ( 16 mixed , 4 C57BL/6 backgrounds ) and five WAP-cre;ApcCKO ( both WAP-cre;ApcCKO/+ and WAP-cre;ApcCKO/CKO ) mammary tumors . We performed 3 separate PCRs to respectively screen for the ( i ) wild-type ( 320 bp ) and ApcCKO/+ ( 430 bp ) alleles , ( ii ) wild-type , ApcCKO/+ and ApcΔ580 ( 500 bp ) alleles , and ( iii ) ApcΔ580 allele alone to genotype tumor DNA as shown in Figure 3A , i–iii . Skin and mammary glands are the tissues known to have transgene expression in K14-cre mice [16] and the presence of the deleted allele of Apc can be detected in small quantities in normal mammary glands of K14-cre; ApcCKO/+ mice only by the ApcΔ580 allele-specific PCR ( Figure 3A iii; G1∼G3 ) . In contrast , most K14-cre; ApcCKO/+ tumors had ApcΔ580/+ genotype , with no or reduced presence of ApcCKO allele ( Figure 3A i and ii; T1∼T3 ) , showing that these tumors were derived from the clonal expansion of ApcΔ580/+ cells . It is important to note that all K14-cre; ApcCKO/+-derived mammary tumors ( 20 of 20 ) were heterozygous for ApcΔ580 mutation but still retained the wild-type allele and did not show allelic loss . In WAP-cre positive ApcCKO mice , the presence of the ApcΔ580 allele was only detected in multiparous WAP-cre;ApcCKO mammary glands ( Figure S1E iii; G1∼3 , G6 , G7 ) but none in mammary glands of nulliparous WAP-cre;ApcCKO females ( Figure S1E iii; G4 , G5 , G8 , G9 ) , confirming the specificity of the WAP promoter . A single prominent ApcΔ580 band was detected by 3-allele screening PCR in four out of five WAP-cre-induced ApcCKO tumors irrespective of parity ( Figure S1E ii; T5 , T6 , T7 , T9 ) , demonstrating the complete conversion of the conditional allele to the deleted allele . One WAP-cre;ApcCKO/+ tumor from a multiparous female was heterozygous for ApcΔ580 mutation with retention of the wild-type allele ( Figure S1E ii; T3 ) while the other from a nulliparous female showed a reduced presence of the wild-type allele , suggesting an allelic loss ( Figure S1E ii; T5 ) . Those tumors that sporadically developed in cre-negative control mice were negative for the ApcΔ580 allele ( Figure S1E iii; T ) , implicating that their development was independent from Cre-induced Apc mutation . These genotyping results were further supported by RT-PCR of corresponding tumor RNA ( data not shown ) . To determine whether the inactivation of the remaining functional Apc allele was achieved by intragenic truncation mutations , we analyzed the tumor DNA by in vitro transcription and translation ( IVTT ) assay . In view of prior mutational analyses in humans and mice , the region of Apc considered most likely to contain mutations is the first 3 kb of exon 15 [19] , [20] . All mammary tumors that showed retention of the wild-type Apc allele were analyzed by IVTT , and truncated Apc products were detected in 19 of 20 ( 95% ) K14-Cre; ApcCKO/+ and all 3 ( 100% ) ApcΔ580/+ mammary tumors , as well as in a single WAP-Cre; ApcCKO/+ mammary tumor that did not show loss of the wild-type allele ( Table 2 ) . One sporadic ApcCKO/+ mammary tumor contained two distinct Apc mutations . The relevant PCR products were subsequently cloned and sequenced . All 25 mutant sequences identified are shown in Table 2 . When histologically distinct portions of a tumor were grossly identifiable as in the tumor in Figure 1C , they were collected separately and were analyzed by IVTT . It was found that they shared the same somatic truncation mutation , further supporting that these histologically distinct tumors derived from a clonal expansion of the same progenitor cell that have acquired an Apc truncation mutation ( Figure 3B , T4&T5 ) . Most of mutations identified in mammary tumors were unique , and were previously not detected in intestinal tumors from Apc1638N/+ heterozygous mice , with or without mismatch repair deficiency [19] , [20] . All but four Apc truncation mutations detected in K14-Cre; ApcCKO/+ mammary tumors were frameshift mutations ( 78 . 9% ) of which two were intragenic deletions of over 10 bp . Most notably , despite the variety of mutant sequences , most of the mutations found were clustered further downstream , beyond codon 1500 , than the mutation cluster region of Apc mutations ( codons 850–1470 ) frequently found in mouse gastrointestinal tumors [19] , [20] ( Figure 3C ) . It is of interest to note that a sporadic ApcCKO/+ mammary tumor contained two distinct Apc mutations that were located in very different regions of the Apc gene . One mutation would result in a truncated product that lacks all the β-catenin binding domains analogous to ApcΔ580 mutation , while the other was located in the same region where mutations in K14-Cre; ApcCKO/+ mammary tumors were found . These results indicate that not only the inactivation of the remaining wild-type allele of Apc is a pre-requisite in these tumors but there is also a selection for particular types of Apc somatic truncation mutations that are likely to result in some retention of down-regulating β-catenin signaling . Wnt1-induced mammary tumors frequently contain activating H-Ras mutations [21] and mutations in K-Ras or N-Ras are frequently found in c-Myc-induced tumors [22] . To determine whether secondary somatic mutations in Ras are involved in Apc mutation-induced mammary tumorigenesis , cDNA copies of tumor Ras mRNAs were analyzed by direct sequencing . In our sequence-based studies of 17 K14-cre; ApcCKO/+ , five WAP-cre; ApcCKO ( ApcCKO/CKO and ApcCKO/+ combined ) , and three ApcΔ580/+ mammary tumors , activating mutations were found at codons 12 and 61 of either H-Ras or K-Ras only in a subset ( 7 of 17 ) of K14-cre; ApcCKO/+ tumors but none from other models . There were four mutations in K-Ras and two in H-Ras and these mutations were mutually exclusive ( Figure 3D , Table 3 ) . It is of interest that , although the incidence is low , both H- and K-Ras activation mutations were found in K14-cre; ApcCKO/+ mammary tumors , since K-Ras mutations were frequently found in Myc-induced mammary tumors but not H-Ras and vice versa in MMTV-Wnt1 tumors . This further supports the molecular diversity as well as the histological heterogeneity of K14-cre; ApcCKO/+ mammary tumors . H-Ras mutations were found in two of four tumors that were predominantly of acinar histology , whereas K-Ras mutations were found in tumors that either predominantly composed of undifferentiated or mixed with undifferentiated histology ( 4 of 8 ) . Sequence-based analysis of exons 5 to 8 of the Tp53 gene was also carried out but Tp53 missense mutations were not detected . There are many mouse models of breast cancer that are different in histopathology and possibly in cell of origin [23] , [24] and these models also have distinct gene expression profiles [25] , [26] . To determine what types of mammary tumor models K14-cre; ApcCKO/+ mice represent , gene expression profiles of three acinar-type mammary tumors , the most frequently found histological pattern in K14-cre; ApcCKO/+ in the mixed background , and three cre-negative normal mammary gland samples were determined using Affymetrix GeneChip M430 2 . 0 arrays . All 3 tumors were heterozygous for ApcΔ580 mutation and the remaining allele contained a truncation mutation in Apc . Both Gene Set Enrichment analysis ( GSEA ) and Ingenuity Pathway analysis ( IPA ) results indicated that mouse tumor profiles have gene sets characteristics of cell cycle , cellular movement and cancer related genes ( Dataset S3 , S4 , S5 , and S6 ) . To get additional insights into K14-cre; ApcCKO/+ tumors , gene expression data from our model was compared to data set of multiple mouse mammary carcinoma models previously published [25] . Based on mouse model intrinsic gene set cluster analysis [25] , a dendrogram and a heatmap were generated using dChip ( http://biosun1 . harvard . edu/complab/dchip/ ) ( Figure 4A–H ) . The gene expression pattern in dendrogram showed more correlations with luminal-type mammary tumors , which include MMTV-Neu , MMTV-PyMT and WAP-Myc [25] . K14-cre; ApcCKO/+ tumors also expressed genes that are strongly expressed in human luminal tumors , such as XBP1 and luminal cell marker K8 and K18 , but were low in basal tumor-defining genes . These tumors also showed high expression of Folate receptor 1 ( Folr1 ) , which is commonly up-regulated in luminal tumor mouse models [25] . As with most mouse mammary tumors , our model was also negative for ER and many estrogen-regulated genes . In agreement with these data , K14-cre; ApcCKO/+ acinar-type mammary tumors expressed K8 while K14 or α-SMA staining was restricted to a myoepithelial pattern ( Figure 2B–D ) . To confirm the initial comparative results , we further compared the gene expression of our K14-cre; ApcCKO/+ model to those previously published for MMTV-Neu , MMTV-Wnt1 and MMTV-Wnt1/Neu bitrangenic mice [26] . A clustering diagram was obtained that clustered tumors from MMTV-Neu , MMTV-Wnt1 and MMTV-Wnt1/Neu bitrangenic samples the same way as indicated previously [26] , with K14-cre; ApcCKO/+ tumors clustering in a separate cluster when hierarchical clustering analysis was done by samples using rank correlation as distance measure for 19 , 581 probes . However if differentially expressed probes from K14-cre; ApcCKO/+ tumors versus controls were used for clustering , K14-cre; ApcCKO/+ tumors clustered next to MMTV-Neu with no effect on clustering between MMTV-Neu , MMTV-Wnt1 and MMTV-Wnt1/Neu samples . The same overall clustering was obtained using mouse model intrinsic gene set [25] ( Figure 4A–H ) . These results support the view that acinar-type mammary tumors from K14-cre; ApcCKO/+ model are luminal type that correlate more with MMTV-Neu than MMTV-Wnt1 model . The elevated expression of Folr1 in mammary tumors detected by microarray analysis was confirmed by quantitative real-time RT-PCR . Since both the IPA and clustering analyses suggested a potential involvement of Neu/Erbb2 ( NM_001003817 ) in K14-cre; ApcCKO/+ mammary tumorigenesis ( Figure S3A ) , we also included Erbb2 in our analysis . Over 30-fold increase in Folr1 ( NM_008034 ) expression compared to the control was detected in acinar-type mammary tumors ( 108 . 7±23 . 3 vs 3 . 1±1 . 1 , p = 0 . 0051 ) confirming the microarray results , but non-acinar type , including those tumors composed primarily of basosquamous , pilar , and undifferentiated structures , and ApcΔ580/+ mammary tumors showed no such differences ( Figure 4I ) . There were no significant differences between tumors and the control for Erbb2 expression .
To delineate the role of Apc mutations in mammary gland , we used Cre-loxP technology to target inactivation of Apc gene in two different mammary epithelial cells , using K14-cre and WAP-cre transgenic mice . The Cre expression in our K14-cre transgenic mice is driven by the basal K14 promoter , which is active in progenitor cells that can give rise to both mammary luminal and myoepithelial lineages [27] , [28] , whereas that of WAP-cre transgenic mice is specific to lactating luminal epithelial cells [17] . The availability of Apc mutant mice under two different mammary promoters and their mammary tumors allowed us to study how Apc loss contributes to mammary tumorigenesis . In this study , we present several lines of evidence that target cells for Apc mutation-induced mammary tumorigenesis are progenitor/stem cells and that they require specific truncation mutations that partially retain β-catenin down-regulating function . First , K14-cre induced ApcΔ580 heterozygosity , but not WAP-cre induced ApcΔ580 heterozygosity or homozygosity , predisposes to mammary tumorigenesis . Second , K14-cre mediated mammary adenocarcinoma showed mixed lineage differentiation , in line with stem or progenitor cell origin , in contrast to WAP-cre mediated tumors that comprised essentially of luminal and abnormal α-SMA positive cells , lacking other basal markers . This is further supported by the fact that two grossly and histologically distinct regions of a tumor share the same somatic Apc truncation mutation , suggesting their origin from a common progenitor . Finally , the remaining wild-type allele of Apc is inactivated not by allelic loss , which is the common mechanism in intestinal tumorigenesis in Apc heterozygous mice , but preferentially by somatic truncation mutations specifically in a well defined region of the gene . This mutation cluster region was different to the one reported for intestinal tumors , implicating that the dosage-specific activation of downstream Wnt/β-catenin signaling pathway is necessary for mammary tumorigenesis . We have previously shown that K14-cre;ApcCKO/CKO mice have aberrant development and squamous metaplasia in many epithelial-derived tissues and die perinatally [14] , not allowing the analysis of Apc loss in postnatal mammary gland . Analogous to K14-driven ApcΔ580 homozygosity , a complete inactivation of the Apc gene in WAP-expressing mammary luminal epithelial cells primarily led to the development of severe squamous metaplasia but rarely neoplasia . These observations suggest that constitutive activation of Wnt/β-catenin signaling pathway by Cre-mediated Apc deficiency , resulting in homozygous Apc ( ApcΔ580/Δ580 ) mutations , invariably induce terminal squamous transdifferentiation of the mammary epithelium irrespective of cell origin of mutated cell , but do not develop tumors . This is in agreement with other Cre-mediated Wnt/β-catenin activation models in which induction of squamous metaplasia but not neoplasia was primarily observed [12] , [13] , [29] . Although homozygous Apc ( ApcΔ580/Δ580 ) mutations induced either by K14 or WAP-promoters invariably results in squamous metaplasia , Cre-mediated Apc heterozygosity ( ApcΔ580/+ ) in K14-cre; ApcCKO/+ mice developed mammary tumors with high penetrance . A similar tendency was also observed in K14-cre; ApcCKO/+ mice backcrossed to C57BL/6 , suggesting that the initiation of Apc-mediated mammary tumorigenesis is not affected by the genetic background . The majority of mammary tumors developed in K14-cre; ApcCKO/+ mice had ApcΔ580/+ genotype and have somatically acquired truncation mutation in the remaining wild-type allele , unlike intestinal tumors in ApcΔ580/+ mice and other germline Apc heterozygotes in which the preferential mechanism of the wild-type Apc is allelic loss [14] , [30]–[32] . Most intriguingly , these truncation mutations were clustered around codon 1530 of Apc ( Figure 3C , Table 2 ) , which is further downstream than the mutation cluster region typically observed in both human and mouse intestinal tumors [19] , [20] , [33] . Regulation of intracellular β-catenin levels is thought to represent one of the most important functions of the Apc tumor suppressor protein . Three different motifs in the central region of Apc are responsible for this activity . The three 15-amino acid ( aa ) repeats that bind β-catenin , the seven 20-aa repeats that both bind and down-regulate β-catenin [34] , [35] , and the three SAMP motifs that bind conductin/axin [36] , [37] . The mutations found in mammary tumors would result in truncated Apc polypeptides retaining up to three of the seven 20-aa repeats but lack all SAMP motifs . It has been shown that loss of these functional motifs , especially those that lead to the elimination of at least five of the seven 20-aa repeats greatly reduced the β-catenin down-regulation activity of Apc [38] . However , Smits et al [39] showed that haploinsufficiency for the truncated Apc polypeptide that retains up to the third 20-aa repeat ( Apc1638N/1572T ) still retained some β-catenin down-regulation activity , resulting in a 5-fold increase in the transcriptional activity compared to a 30-fold increase in Apc mutation homozygosity ( Apc1638N/1638N ) . Since dosage of β-catenin is critical in determining epithelial cell fate in many organs and varying the level of β-catenin signaling during a cell fate program have been shown to switch the epithelial cell fate [40] , it is possible that the homozygosity of Apc mutation ( ApcΔ580/Δ580 ) results in too much β-catenin transcriptional activity , that may push cells into the signaling events that leads to squamous transdifferentiation rather than to hyperplasia of mammary epithelial cells and eventually to neoplasia . The ApcΔ580 heterozygosity with somatic mutation of the remaining wild-type Apc allele that retains some β-catenin down-regulating domains may lead to the optimal dosage of β-catenin necessary for mammary tumorigenesis . Indeed , ApcΔ580/Δ580 homozygosity induced by K14 promoter resulted in severe squmaous metaplasia and ectopic hair follicle morphogenesis in many organs including skin and thymus [14] and the current data using WAP-cre also resulted in squamous metaplasia in mammary glands rather than tumor development . Since mutations found in mouse intestinal tumors result in either allelic loss or a somatic truncation mutation upstream of the third 20-aa repeats [14] , [19] , [20] , [30]–[32] , deletion of all the β-catenin binding domains seems to confer the main selective advantage in mouse intestinal tumorigenesis . Thus , our results and those of others indicate that there is an Apc-regulated level of β-catenin signaling optimal for tumor formation that differs tissue-specifically . The selection for an optimal β-catenin signaling level for tumor formation is also supported by the spectrum of somatic mutations observed in colorectal adenomas from Familial Adenomatous Polyposis ( FAP ) patients with different germline mutations in APC [41] . Our data also partly explain why breast cancers do not develop as frequently as colorectal tumors in FAP patients . It is likely that unless somatic truncation mutation optimal for breast oncogenesis is acquired in the APC gene , mammary epithelial cells are not initiated towards tumorigenesis but instead become metaplastic . Such selection would reduce the incidence of tumorigenesis and require much longer latency , as demonstrated by ApcΔ580/+ and K14-cre; ApcCKO/+ mouse models . Our histopathological and molecular analyses showed that the majority of K14-cre;ApcCKO/+ mammary tumors are adenocarcinomas with multiple foci of squamous metaplasia . These tumors are highly proliferative , ER-negative carcinomas , showing strong positivity for Tcf/β-catenin target genes , Myc and cyclin D1 , with expression of both luminal and basal epithelial markers . They have many histological features common to Wnt pathway tumors previously described [18] , but at the same time have expression profiles that correlate more to luminal tumor models . The latter observation could be explained partly by the predominance of acinar-type histology , which is a luminal type histology , found in the majority of K14-cre;ApcCKO/+ tumors from the mixed background and was the histological type selected for the expression analysis in the current study . The selection of this particular histological type may have biased the expression profiles towards luminal expression pattern . Interestingly , the frequency of the acinar-type histology diminished and basosquamous/pilar structures predominated in the K14-cre;ApcCKO/+-B6 tumors . Since either histological type of mammary tumors had the same mechanism of Apc inactivation irrespective of the genetic backgrounds , it suggests that the mode of initiation is the same but the progression to certain histological types is greatly influenced by modifier genes associated with the genetic backgrounds . Comparison of our gene expression profiling data to published mouse and human breast cancers suggests that acinar-type tumors from K14-cre;ApcCKO/+ mice are more similar to luminal type mouse models [25] . The similarities between mouse and human luminal tumors are limited by the fact that most human luminal epithelial cluster contains the ER and many estrogen-regulated genes , but many mouse mammary tumors , including K14-cre;ApcCKO/+ tumors , are ER-negative . However , the expression profiles of K14-cre;ApcCKO/+ acinar tumors also included a human luminal tumor-defining gene , XBP1 [42] , [43] and stained positive for K8 . Our tumor set also showed elevated expression of Folr1 , which is a gene included in luminal epithelial gene expression cluster that is highly expressed in MMTV-PyMT , MMTV-Neu , and WAP-myc tumors [25] . It was of interest that Folr1 expression level varied within K14-cre;ApcCKO/+ mammary tumors depending on histology , and the ones that had elevated expression were those predominantly composed of acinar histology . The tumors from ApcΔ580 mice , which were of pilar and basosquamous histology , had very low Folr1 expression level . Although Folr1 is not associated with human luminal tumors , its overexpression and poor prognosis have been implicated in human breast cancers [44] , [45] . To determine whether other oncogenic pathways are involved in Apc mutation-induced mammary tumorigenesis , we examined the status of Ras oncogenes and Tp53 in these tumors . It has been previously shown that c-Myc induced mammary tumors in mice frequently harbor spontaneous activating mutations in K-Ras [22] while over 50% of MMTV-Wnt1 tumors contain oncogenic mutations in H-Ras [21] . Jang et al also suggested that K-Ras activation strongly synergizes with both c-Myc and Wnt1 in mammary tumorigenesis and promotes the progression of tumors to oncogene independence , while H-Ras mutant Wnt1-induced tumors remain oncogene dependent [46] . It is of interest that although their presence was mutually exclusive to each other , both H-Ras and K-Ras oncogenic mutations were found in a subset K14-cre;ApcCKO/+ tumors . We could not find a strong association between histology , incidence of lung metastasis and Ras mutations , but those tumors that contained K-Ras mutations frequently had heterogeneous histology containing aggressive looking undifferentiated regions that have lost both lineage markers while those with H-Ras mutations were predominantly of acinar-type . In conclusion , our study demonstrates that activation of Wnt/β-catenin signaling via inactivation of Apc leads to mammary tumorignesis when the inactivation takes place in mammary epithelial progenitor cells rather than more differentiated secretory luminal cells; and when somatic truncation mutation is acquired in a particular region of the Apc gene , which may be necessary to achieve a certain level of β-catenin signaling activation required for mammary tumorigenesis . These initiated tumor cells develop into heterogeneous tumor containing different histological types , each with different expression pattern of lineage markers , some acquiring more oncogenic mutations , such as in either H-Ras or K-ras genes . Our data indicate that only a specific subset of somatic mutations at the Apc gene will successfully lead to tumor formation in the mammary epithelium and there is a selection for Apc mutations that retain some down-regulating activity of β-catenin signaling . We propose that this selection is aimed at a specific level of β-catenin signaling optimal for mammary tumorigenesis , rather than at its constitutive activation achieved by deletion of all the β-catenin down-regulating domains in Apc , which invariably results in squamous metaplasia .
The Apc conditional ( ApcCKO ) and germline ( ApcΔ580 ) knockout mice , WAP-cre and K14-cre transgenic mice have previously been described [14] , [16] , [17] . ApcΔ580/+ mice have already been backcrossed to C57BL/6 background . The rest of the mice analyzed in this study were generated as follows: ApcCKO heterozygote mice of the F1 generation ( C57BL/6×129/S background ) were first crossed with either WAP-cre mice ( C57BL/6 background ) or K14-cre transgenic mice ( FVB background ) . Cre-positive ApcCKO/+ male mice thus generated were then crossed with ApcCKO/CKO females to generate homozygous and heterozygous ApcCKO offspring either with or without respective cre transgene . The mice were intercrossed thereafter for maintenance . Subsequently , K14-cre; ApcCKO/+ female mice backcrossed for 4 generations to achieve >95% C57BL/6J background were also included in the analysis so that the results between two different promoters will be comparable . Females with genotypes WAP-cre; ApcCKO/+ and WAP-cre; ApcCKO/CKO were mated and undergone pregnancies 4 times to facilitate WAP-cre-mediated deletion of exon 14 of Apc gene . Mice with genotype K14-cre; ApcCKO/CKO are perinatally lethal [14] and phenotypically normal K14-cre; ApcCKO/+ littermates were used for analysis . The mice were sacrificed when either they were moribund or their tumors reached at least 2 cm in diameter , following Institutional Animal Care and Use Committee guidelines . Mouse tails tips obtained at ∼10 days of age were lysed overnight in DirectPCR Lysis Reagent ( Viagen Biotech ) containing 0 . 1 mg/ml Proteinase K ( Qiagen ) . The crude lysates were incubated at 85°C for 45 minutes and 0 . 5 µl of lysate was directly used per 25 µl PCR reaction . Detection of various Apc alleles and cre transgene was carried out as previously described [14] . Mice were sacrificed by CO2 inhalation when they developed gross tumors or were moribund . The location and size of tumors were routinely recorded and pictures were taken from each mouse . Tumors were cut in portions and a portion was either fixed in 10% netural buffered formalin ( NBF ) or in 4% paraformaldehyde . The other portion was either snapped frozen in liquid nitrogen or immersed in RNAlater solution ( ambion ) overnight and stored at −80°C until molecular analyses . Mammary glands were also collected routinely from each mouse: 4th mammary gland was fixed flat on a piece of paper towel in 10% NBF , 9th mammary gland for whole mount and either 8th and 3rd tumor-free mammary gland was collected for molecular analyses . The mice were then dissected for gross examination . A portion of liver and lungs were similarly collected and fixed . Then the whole body was fixed in Bouin's solution . The fixed tissue samples were then submitted to Rodent Histopathological Core , processed and embedded in paraffin . Tissue sections were cut and stained with H&E for histopathological examinations . Five-µm sections were cut from the paraffin-embedded tissues and immunohisochemistry was performed essentially as previously described [14] . Briefly , sections were deparaffinized , rehydrated and boiled in either Citrate buffer ( 10 mM , pH 6 ) or Tris buffer ( 10 mM Tris , 1 mM EDTA , pH 9 ) for antigen retrieval . Slides were than treated with 3% peroxidase in PBS , followed by blocking in normal horse serum . Primary antibodies against Ki67 ( 1∶200 , Vector Laboratories ) , β-catenin ( 1∶200 , BD Transduction Lab ) , cyclin D1 ( 1∶100 , Lab Vision ) , c-myc ( 1∶200 , Upstate ) , esterogen receptor ERα and progesterone receptor ( 1∶2000 , 1∶500 , respectively , Santa Cruz ) and cellular markers such as cytokeratins K1 , K6 , K14 ( 1∶1000 , 1∶500 , 1∶2000 , respectively , Covance ) , K8 ( 1∶100 , TROMA-I , DSHB ) , p63 ( 1∶200 , Chemicon International ) and α-smooth muscle actin ( α-SMA , 1∶800 , Sigma ) were applied followed by an incubation with biotin-conjugated appropriate secondary antibody . Mouse-on-Mouse kit ( Vector Laboratories ) was used with the mouse primary antibodies . The Vectastain Elite ABC kit and DAB ( Vector Laboratories ) were used for detection , following manufacturer's instructions . This procedure was carried out as described on the mammary gland website: http://mammary . nih . gov . Both genomic DNA and RNA from tumors and various tissue samples collected at the time of autopsy were extracted as described previously [14] . Briefly , genomic DNA was extracted using DNeasy mini kit ( QIAGEN ) . RNA was extracted by homogenizing tumors and tissues in 3 ml Trizol reagent ( Invitrogen ) . After phase separation , an equal volume of 70% ethanol was added to the aqueous phase and purified through PureLink Micro-to-Midi Total RNA Purification System ( Invitrogen ) , following manufacturer's instruction . Concentrations of nucleic acids were determined by Nanodrop ( Ribogreen , Molecular Probes , Eugene , Oregon , United States ) . Tissue-specific recombination of the conditional alleles in tumors and various tissue samples was examined by analyzing both their extracted DNA and RNA as described previously [14] . Briefly , genomic DNA samples were examined by a semi-quantitative 3-primer genotyping PCR that will give 3 distinct sized products from the wild-type ( 320 bp ) , conditional ( 430 bp ) and deleted ( 500 bp ) alleles of Apc . In addition , two separate 2-primer PCRs were performed; one to check for the wild-type and ApcCKO alleles and the other to detect the presence of the ApcΔ580 allele alone . The expression of full-length and truncated Apc alleles were examined by performing RT-PCR on RNA using SuperScript One-Step RT-PCR with Platinum Taq ( Invitrogen ) , following manufacturer's protocol . Upon confirmation of the retention of the wild-type Apc allele , codons 677–1674 of the mouse Apc gene were analyzed for truncation mutations by PCR and in vitro transcription and translation ( IVTT ) assay as described previously [20] , [47] but with some modifications . All DNA amplifications were performed using Pfu Ultra II fusion HS DNA polymerase ( Stratagene ) according to manufacturer's instruction . The wild-type allele-specific amplification of Apc was performed by nested-PCR to eliminate co-amplification of deleted ApcΔ580 allele using the forward primer 5′-CATTCTCCCCTACTTAGATGG and a reverse primer 5′-GTTGTCATCCAGGTCTGGTG in the first PCR reaction . Two overlapping segments of the Apc gene covering codons 677–1234 and 1100–1674 were subsequently amplified from aliquots of the first reactions using two pairs of PCR primers specific for IVTT . Cycling conditions for the first stage PCR were one cycle of 94°C for 2 min , followed by 20 cycles of 94°C for 20 s , 58°C for 20 s and 72°C for 90 s , with one final extension cycle at 72°C for 3 min . Cycling conditions for the second stage PCR were as above except the cycle number was increased to 25 . The PCR products were directly used for IVTT assay as described previously [20] , [47] . In order to facilitate the detection of Apc truncation mutations from tumor samples that are frequently co-harvested with adjacent normal tissues , we developed a method based on expression of Apc-GFP fusion polypeptides in bacteria , in which colonies derived from PCR products with mutation appear GFP-negative . The pTrcHis B Prokaryotic Expression Vector ( Invitrogen ) was modified to contain GFP coding sequence ( Acc# U87625 ) between NheI and SacI sites . The sequence between NcoI and NheI sites , coding poly-histidine region was replaced with a BamHI site , such that insertion of amplified Apc fragments between BamHI and NheI sites restores the reading frame of GFP . For characterization of tumor-specific mutations , the PCR products were digested with BamHI and NheI , gel purified , and cloned into modified pTrc vector and transformed into bacterial cells using standard cloning procedures . Transformed bacterial cells were spread on LB plates containing final concentrations of 50 µg/ml ampicillin and 50 µM IPTG and incubated overnight at 37°C . The numbers of non-fluorescent and fluorescent colonies were counted under long-wave length UV light . When the percentage of non-fluorescent colonies over total was above the control level , individual GFP negative clones were screened by IVTT to identify mutations , and their DNA sequences were determined . For p53 mutational analysis , exons 5 to 8 of Tp53 were examined by performing RT-PCR on RNA using primers p53-F117 and p53-R313 and SuperScript One-Step RT-PCR with Platinum Taq ( Invitrogen ) . The PCR products were then sequenced by nested primers , p53-F121 and p53-R308 . For K-Ras and H-Ras codons 12 , 13 or 61 mutation analysis , RNA was first reverse-transcribed and amplified using SuperScript One-Step RT-PCR system , followed by nested-PCR using Pfu Ultra II fusion HS DNA polymerase ( Stratagene ) . The mutations were scored positive when approximately half of the resulting amplified DNA had the same mutation , and when the results were confirmed by either sequencing from both ends or by sequencing corresponding genomic amplified products . Sequences of primers used in the study are listed in Table S1 . Total RNA ( 5–10 µg ) extracted from 3 acinar-type mammary tumors and 3 age-matched control mammary glands from cre-negative mice were hybridized and scanned to GeneChip M430 2 . 0 according to Affymetrix protocols ( Affymetrix ) . Scanned microarray images were imported into GeneChip Operating Software ( GCOS , Affymetrix ) to generate signal values and absent/present calls for each probe-set using the MAS 5 . 0 statistical expression algorithm ( . chp files ) . Each array was scaled to a target signal of 500 using all probe-sets and default analysis parameters . Prior to performing any other analysis Affymetrix detection calls were used to remove 12 , 844 probes which had ‘Absent’ call across all samples . Data set with 32 , 322 probes was used as starting point for any subsequent analysis . To identify genes differentially expressed between tumor and control samples two-sample t-test was used ( Dataset S1 ) . T-test was performed using ComparativeMarkerSelection module of Gene Pattern ( http://www . broad . mit . edu/cancer/software/genepattern/ ) [48] . Based on two-sample t-test results , any probes with fold differences below 2 , t-test values below 4 . 5 were removed and only the probes that have either consistent absent or present calls were used as input for Ingenuity Pathway Analysis ( IPA ) Software ( http://www . ingenuity . com/ , Dataset S2 ) . Ingenuity core analysis generated over 60 networks with nearly 600 network nodes in total , many of them involved in cell cycle , cell , growth , cell death , DNA replication , and cancer ( Dataset S3 , S4 , S5 , and S6 ) . Gene Set Enrichment Analysis ( GSEA ) was performed using java GSEA http://www . broad . mit . edu/gsea/ as previously described [49] To get additional insights in K14-cre; ApcCKO/+ tumors , gene expression from our mouse model was compared to data set of multiple mouse mammary carcinoma models and human breast tumors previously published [25] . In particular , based on genes mainly in intrinsic gene set cluster analysis of mouse models a dendrogram was generated using dChip ( http://biosun1 . harvard . edu/complab/dchip/ ) ( Figure 4A ) . To further compare gene expression of K14-cre; ApcCKO/+ model , gene expression probe level data ( CEL files 430A 2 . 0 ) from Shixia Huang et al . for tumors from MMTV-Wnt1 , MMTV-Neu and MMTV-Wnt1/MMTV-Neu bitransgenic mice were obtained [26] . MAS5 . 0 algorithm was used to estimate probe expression . Since 430 2 . 0 chips have additional probe sets compared to 430A 2 . 0 the subset corresponding to 430A 2 . 0 was used for MAS5 . 0 global method of scaling/normalization . Target Intensity value of 500 was used for all arrays . Prior to performing any other analysis , Affymetrix detection calls were used to remove 3 , 045 probes which had ‘Absent’ call across all samples . Data set with 19 , 581 probes was used as starting point for any subsequent analysis . Hierarchical clustering by samples was performed using data for all 19 , 581 probes as input and rank correlation as ( Figure S3B ) . Subsequently , 1 , 335 differentially expressed probes ( fold differences above 2 , t-test values above 4 . 5 ) from K14-cre; ApcCKO/+ tumors versus controls were used for clustering analysis ( Figure S3C ) . The expression levels of Folr1 and Erbb2 in 17 K14-cre; ApcCKO/+ , 5 WAP-cre; ApcCKO ( ApcCKO/CKO and ApcCKO/+ combined ) , 4 ApcΔ580/+ mammary tumors were compared to those of 7 normal mammary glands by quantitative RT-PCR with HPRT1 as an internal control . Most of K14-cre; ApcCKO/+ mammary tumors in the mixed background consisted of acinar-type , but there were a few which had distinct histology . Therefore , tumors were roughly divided into 2 groups; acinar and non-acinar . The latter group of mammary tumors primarily composed of undifferentiated , basosquamous and pilar-type histology . We analyzed 14 acinar-type , and 8 non-acinar type . The mammary tumors developed in ApcΔ580/+ mice were mostly of basosquamous-type . TaqMan Gene Expression Assays for respective genes were used on 7500 Fast Real-Time PCR System ( Applied Biosystem ) according to the manufacturer's protocol . Relative quantity was calculated using Sequence Detection Software version 1 . 4 ( Applied Biosystem ) . | Breast cancer is one of the most common malignanices in women in Western countries . Many components of Wnt/β-catenin signaling pathway are known to play critical roles in mammary tumor development , yet the role of the tumor suppressor gene APC ( adenomatous polyposis coli ) in breast oncongenesis is unclear . To study the role of Apc in mammary tumorigenesis , we introduced conditional Apc mutations specifically into two different mammary epithelial populations using K14 ( Keratin 14 ) -cre and WAP ( Whey Acidic Protein ) -cre transgenic mice that express Cre recombinase in mammary progenitor cells and lactating luminal cells , respectively . In this study , we show that a specific type of Apc somatic mutations in mammary progenitor/stem cell population in mice induces mammary carcinomas with histological and molecular heterogeneity , but a complete deletion leads to squamous metaplasia . Our data show that mutations in a multilineage progenitor cell are important in certain mammary tumors . We also show that certain Apc mutations are selected to achieve a specific level of β-catenin signaling optimal for mammary tumor development and explain partially the reason why breast cancers do not develop as frequently as colorectal tumors in patients that carry germline mutations in APC . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/disease",
"models",
"genetics",
"and",
"genomics/cancer",
"genetics",
"oncology/breast",
"cancer"
] | 2009 | Genetic Mechanisms in Apc-Mediated Mammary Tumorigenesis |
Vegetative phase change is regulated by a decrease in the abundance of the miRNAs , miR156 and miR157 , and the resulting increase in the expression of their targets , SQUAMOSA PROMOTER BINDING PROTEIN-LIKE ( SPL ) transcription factors . To determine how miR156/miR157 specify the quantitative and qualitative changes in leaf morphology that occur during vegetative phase change , we measured their abundance in successive leaves and characterized the phenotype of mutations in different MIR156 and MIR157 genes . miR156/miR157 decline rapidly between leaf 1&2 and leaf 3 and decrease more slowly after this point . The amount of miR156/miR157 in leaves 1&2 greatly exceeds the threshold required to specify their identity . Subsequent leaves have relatively low levels of miR156/miR157 and are sensitive to small changes in their abundance . In these later-formed leaves , the amount of miR156/miR157 is close to the threshold required to specify juvenile vs . adult identity; a relatively small decrease in the abundance of miR156/157 in these leaves produces a disproportionately large increase in SPL proteins and a significant change in leaf morphology . miR157 is more abundant than miR156 but has a smaller effect on shoot morphology and SPL gene expression than miR156 . This may be attributable to the inefficiency with which miR157 is loaded onto AGO1 , as well as to the presence of an extra nucleotide at the 5' end of miR157 that is mis-paired in the miR157:SPL13 duplex . miR156 represses different targets by different mechanisms: it regulates SPL9 by a combination of transcript cleavage and translational repression and regulates SPL13 primarily by translational repression . Our results offer a molecular explanation for the changes in leaf morphology that occur during shoot development in Arabidopsis and provide new insights into the mechanism by which miR156 and miR157 regulate gene expression .
Leaves produced at different times in shoot development are often morphologically distinct . In Arabidopsis , for example , successive rosette leaves differ in size , length:width ratio , the angle of the leaf base , hydathode number , the complexity of the vascular system , cell size , sensitivity to gibberellic acid , and the absence vs . the presence trichomes on the abaxial surface of the leaf blade [1–7] . Some of these so-called "heteroblastic" traits change gradually throughout shoot development , others change early in shoot development and are then expressed more-or-less uniformly , while still others are present at one stage of development and absent at a different stage . These latter two patterns allow the shoot to be divided into several discrete phases , the transition between which is termed "vegetative phase change" [8] . miR156 is the master regulator of vegetative phase change in Arabidopsis [3 , 9] and other flowering plants [10–15] . It is initially expressed at a very high level , and declines as the shoot develops [3 , 9 , 16–18] . This decrease is associated with an increase in the expression of its targets , SQUAMOSA PROMOTOR BINDING PROTEIN-LIKE ( SPL ) transcription factors , and is responsible for the transition to the adult phase . Most species also possess another miRNA , miR157 , that differs from miR156 at 3 nucleotides [19] . miR157 has the same targets as miR156 and produces an over-expression phenotype similar to that of miR156 [20] . However , the normal function of miR157 is still unknown . Although it is clear that miR156 , and possibly miR157 , regulate many of the changes that occur during shoot development , the function of these miRNAs at specific times in development and in specific leaves is poorly understood . In particular , it remains to be determined if miR156 is responsible for the graded changes in leaf morphology that occur during the juvenile phase and , if so , how it produces this variation . It is also important to determine if miR156 plays a role in shoot morphogenesis during the adult phase . Although miR156 is present at lower levels in the adult phase than in the juvenile phase , a comparison of the expression patterns of miR156-sensitive and miR156-insensitive SPL reporters suggests that miR156 represses SPL gene expression during both phases , albeit to different extents [21] . Finally , it is important to determine the mechanism by which miR156 represses gene expression . Previous studies have shown that miR156—as well as several other plant miRNAs ( reviewed in [22] ) —mediates both transcript cleavage and translational repression [23–27] , but the relative importance of these processes for the activity of miR156 remains to be determined . This question is of particular interest in light of the observation most SPL transcripts change very little during shoot development , despite the significant decrease in miR156 that occurs during this process [21] . To address these questions , we characterized the morphological and molecular phenotype of loss-of-function mutations in MIR156 and MIR157 genes , and measured the absolute amount of miR156/miR157 in successive leaf primordia . We also quantified the effect of varying miR156/miR157 levels on the expression of their SPL targets . Our results demonstrate that miR156 and miR157 have different expression patterns , different activity , and mediate transcript cleavage and translational repression to different extents at different SPL genes . We also show that variation in the level of miR156/miR157 only has a significant effect on SPL gene expression when these miRNAs are present at relatively low levels . These results provide a foundation for detailed studies of the molecular mechanism of miR156/miR157 activity and their role in shoot morphogenesis .
In Arabidopsis , miR156 is encoded by 8 genes and miR157 is encoded by 4 genes . We characterized the contributions of these genes to the overall pool of miR156/miR157 by sequencing small RNAs from the FRI FLC and FRI flc-3 genotypes [28 , 29] . We chose these genotypes because they represent common genotypes in naturally-occurring accessions of Arabidopsis [30] , and because the vegetative and flowering phenotype of FRI flc-3 is nearly identical to that of wild-type Col [2 , 31] . Sequencing of small RNAs from 11-day-old shoot apices ( 2 replicates of each genotype ) revealed an abundant 20 nt transcript that maps to MIR156A , B , C , D , E , and F , an abundant 21 nt transcript that maps to MIR157A , B , and C , a moderately abundant 21 nt transcript that maps to MIR156D , and 3 rare transcripts that map uniquely to MIR156G , MIR157D and MIR156H ( Table 1 ) . Unexpectedly , miR157-related transcripts were more abundant than miR56-related transcripts . To determine which genes produce these transcripts , we identified T-DNA insertions in MIR156A , MIR156C , MIR156D , MIR157A , and MIR157C , and used site-directed mutagenesis to produce mutations in MIR156B . RT-qPCR analysis of these alleles demonstrated that they eliminate or greatly reduce the primary transcripts of the affected genes ( S1 Fig ) . We then examined the amount of miR156 and miR157 in these stocks by hybridizing RNA blots with probes for miR156 , miR157 , and a combination of both probes . We used this approach instead of RNA sequencing because libraries constructed with two different RNA adaptors revealed that different miRNAs ligate with different efficiencies to each adaptor [32] . Although the miR156 and miR157 probes cross-hybridize to some extent , the source of the hybridization signal could by determined by comparing the effect of mir156 and mir157 mutations on these signals . The effect of mir156 and miR157 mutations on the levels of miR156 and miR157 in 11-day-old seedlings and in 1mm primordia of leaves 1 & 2 is shown in Fig 1 . In Col , the miR156 probe hybridized to 20 nt and 21 nt transcripts , with the 20 nt transcripts being more abundant than the 21 nt transcripts ( Fig 1A ) . The abundance of the 20 nt transcripts was reduced to 62 ± 10% ( ± SD , n = 4 ) of wild-type in mir156a-2 ( hereafter , mir156a ) , to 51 ± 8% ( ± SD , n = 4 ) of wild-type in mir156c-1 ( hereafter , mir156c ) , and to 11 ± 1% ( ± SD , n = 3 ) of wild-type in the mir156a/c double mutant ( Fig 1A ) . These genes are therefore the major source of the 20 nt miR156 transcripts . mir156d-1 ( hereafter , mir156d ) had very little effect on the overall abundance of miR156 in 11 day-old seedlings and leaf primordia ( Fig 1 ) . However , the intensity of the 21 nt band was slightly reduced in mir156d-1 and in genotypes containing this mutation; for example , the 21 nt miR156-hybridizing band was slightly less intense in the mir156a/c/d mir157a/c pentuple mutant than in the mir156a/c mir157a/c quadruple mutant ( Fig 1A ) . MIR156B also makes a minor contribution to the miR156 pool because the intensity of the miR156-hybridizing bands was essentially identical in the mir156a/b/c/d and mir156a/c mutants ( Fig 1A ) , and there was no detectable difference between the intensity of the 20 nt and 21 nt bands in leaf primordia ( LP ) of the mir156a and mir156a/b mutants ( Fig 1B ) . In wild-type Col , the miR157 probe hybridized strongly to 21 nt transcripts and more weakly to 20 nt transcripts ( Fig 1A ) . The 20 nt band was absent in mir156a/c , and thus represents cross hybridization of the miR157 probe with miR156 . The intensity of the 21 nt miR157-hybridizing band was reduced to 81 ± 13% ( ± SD , n = 4 ) of wild-type in mir157a-1 , to 26 ± 5% ( ± SD , n = 7 ) of wild-type in mir157c-1 ( mir157c ) , and to 13 ± 2% ( ± SD , n = 3 ) of wild-type in the mir157a/c double mutant . The remaining 21 nt signal in the mir157a/c double mutant partly reflects cross-hybridization of the miR157 probe with 21 nt miR156 transcripts because the intensity of this band was slightly reduced in the mir156a/c mir157a/c and mir156a/c/d mir157a/c mutants compared to mir157a/c . These results demonstrate that the major miR157 transcript is 21 nt , and that MIR157C is the major source of this transcript . Hybridization with a 1:1 mixture of miR156/miR157 probes revealed that 21 nt transcripts are significantly more abundant than 20 nt transcripts in 11 day-old seedlings and in the primordia of leaves 1&2 ( Fig 1A and 1B ) . The 21 nt band was reduced significantly in mir157a/c , and therefore corresponds primarily to miR157 , whereas the 20 nt band was nearly absent in mir156a/c , and therefore corresponds to miR156 . These results are consistent with the results of RNA sequencing ( Table 1 ) , and demonstrate that miR157 is more abundant than miR156 in young seedlings . Northern analysis using a mixed miR156/miR157 probe revealed that the amount of miR156 and miR157 in the mir156a/c/d mir157a/c pentuple mutant is about 10% of the wild-type level ( Fig 1A and 1B ) . Assuming that the mutations present in this pentuple mutant are null alleles , the amount of miR156/miR157 in this line represents the combined output of MIR156E , F , G , H and MIR157B , D . These six genes therefore contribute relatively little to the production of miR156 and miR157 in seedlings . The morphology of rosette leaves changes qualitatively and quantitatively during shoot development . In plants grown in SD to delay flowering , the first two rosette leaves are small and round , and lack serrations and abaxial trichomes [1 , 2] ( Fig 2A ) . Leaves 3 and 4 are larger than leaves 1 and 2 , but also have round leaf blades with no serrations and no abaxial trichomes . Leaves 5 through 9 are larger , more elongated , and more serrated than the first four leaves . Depending on light quantity and quality , abaxial trichome production begins between leaf 7 and 9 , and is accompanied by a decrease in the angle of the leaf base and by the production of more prominent serrations ( Fig 2B ) . Previous studies have shown that the juvenile forms of these traits require the activity of miR156/miR157 [9] , but the relationship between the abundance of these miRNAs and the changes in leaf morphology that occur during shoot development is still unknown . To begin to answer this question , we measured the abundance of miR156 and miR157 in successive rosette leaves of wild-type plants , and characterized the effect of mir156 and mir157 mutations on leaf morphology . RT-qPCR ( S2A Fig ) and Northern analysis ( S2B Fig ) demonstrate that miR156 and miR157 increase as leaves expand . However , the expression pattern of these miRNAs in successive fully expanded leaves ( S2C and S2D Fig ) and 1 mm LP ( Fig 2 ) is quite similar , indicating that the factors responsible for variation in the expression of miR156/miR157 during shoot development operate at all stages of leaf development . Consistent with our previous analyses of shoot apices [21] , miR156 and miR157 decrease significantly from LP1&2 to LP3&4 , and then decline more gradually before reaching a relatively constant level around leaf 13 ( Fig 2 ) . LP3&4 had approximately 25% , LP9 had 12% , and LP13 had 8% of the amount of miR156 present in LP1&2 . miR157 declined to a lesser extent: LP 3&4 had approximately 50% , LP9 had 25% , and LP13 had 17% of the amount of miR157 present in LP1&2 ( Fig 2 ) . The expression pattern of miR156 in fully-expanded ( FE ) leaves matched its expression pattern in LP , but miR157 did not decline as dramatically between FE1&2 and FE3&4 as it did between LP1&2 and LP3&4 ( S2C and S2D Fig ) . We then determined the absolute amount of these miRNAs in LP by comparing the RT-qPCR results obtained with leaf samples to the results obtained using known quantities of miR156 and miR157 . Synthetic miR156 and miR157 transcripts were serially diluted in 600ng/μl E . coli RNA , and a standard curve was produced by plotting the concentrations of these miR156 and miR157 standards against 2-ct of the corresponding RT-qPCR reaction . RT reactions were performed in parallel using 600ng of total RNA from LP1&2 . The 2-ct value of the LP1&2 sample was then fitted to the standard curve , and the concentration of miR156 or miR157 was calculated using linear regression . This information , and the results of the experiment shown in Fig 2A , were then used to calculate the absolute amount of miR156 and miR157 in other LP ( Fig 2B ) . miR156 was present in LP1&2 at a concentration of 1 . 96 ± 0 . 1 x 105 copies per ng total RNA , whereas miR157 was present at a concentration of 2 . 45 ± 0 . 2 x 105 copies per ng total RNA ( Fig 2B ) . miR156 subsequently declined to approximately 2 . 6 x 104 copies per ng total RNA in LP9 , whereas miR157 declined to 6 . 1 x 104 copies per ng total RNA . Thus , the transition between leaves 1&2 and leaves 3&4 is accompanied by a major decline in the level of miR156 and miR157 whereas subsequent changes in leaf morphology are associated with much smaller changes in the abundance these miRNAs . The juvenile-to-adult transition occurred during the period when miR156 and miR157 were declining very gradually , and was accompanied by a relatively small change in the abundance of these transcripts . The relative importance of different MIR156 and MIR157 genes in shoot development was determined by characterizing the morphological phenotype of plants singly or multiply mutant for mir156a , mir156b , mir156c , mir156d , mir157a and mir157c ( Figs 3 and 4 ) . Plants were grown in SD to eliminate the effect of floral induction on leaf morphology [2] . We measured two traits that change with leaf position—the production of trichomes on the abaxial surface of the leaf blade and the angle of the leaf base . In wild type plants , the angle of leaf base became more acute starting with leaf 5 , and abaxial trichome production started at leaf 9 ( Figs 3 and 4 ) . The effect of mir156 and mir157 mutations on abaxial trichome production and leaf shape was correlated with abundance of miR156/miR157 in different leaves . Leaves produced early in shoot development , which have a relatively high level of miR156/miR157 , were less sensitive to these mutations than leaves produced later in shoot development , which have a relatively low level of miR156/miR157 ( Figs 3 and 4 ) . For example , mir156a , mir157a , mir157c , and mir157a/c caused abaxial trichomes to be produced on leaves 7 and/or 8 , but did not affect abaxial trichome production or the shape of leaf 1 , 3 , and 5 . mir156c produced abaxial trichomes on leaves 7 and 8 and significantly reduced the angle of the leaf base in leaves 3 and 5 , but had no effect on leaf 1 . miR156a/c and mir156a/b/c/d reduced the angle of the leaf blade in leaves 1 , 3 and 5 , but had a more significant effect on leaves 3 and 5 than on leaf 1; these genotypes only produced abaxial trichomes on leaves 6 and above . miR156a/c mir157a/c and mir156a/c/d mir157a/c had a significant effect on the shape of leaves 1 , 3 , and 5 , but rarely produced abaxial trichomes on leaf 2 , and never produced abaxial trichomes on leaf 1 ( Figs 3 and 4 ) . The absence of abaxial trichomes on leaves 1 and 2 is attributable to the small amount of miR156/miR157 remaining in these mutants because 35S::MIM156 consistently produced abaxial trichomes on both of these leaves ( Fig 4A ) . These results demonstrate that abaxial trichome production is more sensitive to miR156/miR157 than leaf morphology , and is strongly repressed by even low levels of these miRNAs . They also reveal that the amount of miR156/miR157 in leaves 1&2 far exceeds the amount required to specify their identity . Only genotypes with very low levels of miR156/miR157 ( e . g . , miR156a/c mir157a/c , mir156a/c/d mir157a/c , 35S::MIM156 ) cause these leaves to resemble adult leaves ( Figs 3B and 4B ) . In general , the morphological phenotype of mir156/mir157 mutations was correlated with their effect on the abundance of miR156 or miR157 . mir156a and mir156c have a relatively large effect on the level of mir156 ( Fig 1 ) and also have a relatively large effect on shoot morphology . mir156c has a more significant effect on the morphology of leaves 3 and 5 than mir156a ( Fig 4B ) , which is consistent with its slightly larger effect on the abundance of miR156 ( Fig 1A ) . mir156b and mir156d have very minor effects on the abundance of mir156 ( Fig 1 ) and also have minor effects on shoot morphology; mir156b did not significantly enhance the phenotype of mir156a or mir156a/c , and mir156d only produced a significant effect on leaf morphology in combination with mir156a/c and mir157a/c . The only unexpected result was the phenotype of mir157a/c . miR157 is more abundant than miR156 and was therefore expected to play a larger role in vegetative phase change than miR156 . However , mir157a/c had a significantly weaker effect on abaxial trichome production and leaf shape than mir156a/c ( Figs 3 and 4 ) , even though these double mutants have approximately the same amount of miR157 and miR156 , respectively ( Fig 1 ) . This observation demonstrates that miR157 is less important for vegetative phase change than miR156 , and suggests that it may be less active than miR156 . miRNAs with a 5’ terminal uridine , such as miR156 and miR157 , repress the expression of their targets via their association with AGO1 [33] . To determine if miR156 and miR157 are loaded onto AGO1 with different efficiencies , we measured the amount of miR156 and miR157 associated with AGO1 in planta . For this purpose , we took advantage of an ago1-36 line transformed with AGO1-FLAG [34] . Extracts from 2-week-old seedlings of this transgenic line and wild-type Col ( as negative control ) were treated with an antibody to the FLAG epitope , and small RNAs were extracted from immunoprecipitad ( IP ) and non-IP samples and assayed using Northern blots . Hybridization with a mixed miR156/miR157 probe revealed that miR157 ( 21 nt band ) was more abundant than miR156 ( 20 nt band ) in the input fraction , but that miR156 was as abundant as miR157 in the IP fraction ( Fig 5 ) . This result indicates that miR156 is more efficiently loaded onto AGO1 than miR157 . This cannot be the only reason for the difference in the phenotypes of mir156a/c and mir157a/c because the amount of miR156 and miR157 associated with AGO1 is quite similar . Another possibility is that the AGO1-miR157 complex is inherently less active than the AGO1-miR156 complex . miR156 and miR157 bind to the SPL2 , SPL9 , SPL10 , SPL11 , and SPL15 transcripts with only one mismatched nucleotide , although the position of this nucleotide is different for the two miRNAs ( Fig 6A ) . In addition to two internal nucleotides , miR157 differs from miR156 in possessing an additional U at its 5' end . This 5’ U is unpaired in the miR157-SPL13 duplex ( Table 1 ) . To determine if this extra nucleotide might influence the activity of miR157 we compared the relative strengths of miR156a and miR156d . The miR156d transcript is identical to the miR156a transcript , except for the presence of an additional 5’U ( Fig 6A ) . The phenotypes of 5 transgenic lines constitutively expressing a genomic fragment containing MIR156A under the regulation of the CaMV 35S promoter , and an equal number of lines containing a similar construct encoding MIR156D [3] , were compared under LD conditions . The lines used for this analysis were selected because they possessed a single T-DNA insertion site . The 35S::MIR156A lines produced approximately twice as many leaves without abaxial trichomes and approximately twice as many cauline leaves as the lines transformed with 35S::MIR156D ( Fig 6B ) . This result demonstrates that MIR156D is less effective than MIR156A , and suggests that the additional 5' U in miR157 is partly responsible for its lower biological activity . To determine the molecular basis for the effect of mir156 and mir157 mutations on leaf morphology , we compared SPL transcript levels in LP1&2 and LP3&4 in wild-type and mir156/mir157 mutant plants ( Fig 7 ) . Consistent with their modest effect on leaf morphology , single mir156c and mir157c mutations had a very small effect on SPL transcripts . However , plants with multiple mir156 and/or mir157 mutations displayed a significant increase in the level of some SPL transcripts . SPL3 transcripts were particularly responsive to a decrease in the level of miR156 , increasing about 4-fold in mir156c and 5-to-6-fold in mir156a/c . In contrast , SPL3 transcripts were relatively insensitive to a decrease in miR157 , except in genotypes that were also deficient for miR156 . For example , SPL3 was elevated nearly 20 fold in LP3&4 of the mir156a/c/d mir157a/c pentuple mutant . SPL9 and SPL15 transcripts increased very slightly in mir156a/c and mir157a/c but increased up to 6-fold in mir156a/c mir157a/c and mir156a/c/d mir157a/c . SPL2 , SPL10 and SPL11 increased 2-fold or less in mir156a/c and mir157a/c , and only about 3-fold in mir156a/c mir157a/c and mir156a/c/d mir157a/c . SPL13 transcripts were unaffected in mir157a/c , were elevated about 2-fold in both mir156a/c and mi156a/c mir157a/c , and were only slightly more abundant than this in mir156a/c/d mir157a/c . The abundance of SPL3 is regulated directly by miR156/miR157 via miRNA-induced transcript cleavage , and indirectly by the effect of miR156/miR157-regulated SPL proteins on the expression of miR172 , which in turn represses a group of AP2-like genes that repress the transcription of SPL3 , SPL4 , and SPL5 [3 , 9 , 23 , 35] . This combination of direct post-transcriptional regulation by miR156/miR157 and indirect transcriptional regulation via the miR172-AP2 pathway may be responsible for the hypersensitivity of SPL3 transcripts to variation in the abundance of miR156/miR157 . In contrast , the only way in which miR156/miR157 have been found to regulate the expression of other SPL genes is through a direct interaction with their transcripts . These findings therefore suggest that the SPL2 , SPL9 , SPL10 , SPL11 , SPL13 and SPL15 transcripts are differentially sensitive to destabilization by miR156 and miR157 . Both the transcripts and the protein products of miR156/miR157-regulated SPL genes increase during shoot development [3 , 18 , 21 , 36 , 37] . The expression patterns of miR156/miR157-resistant reporter genes suggest that this increase is largely mediated by miR156/miR157 [21] , but whether miR156/miR157 are entirely responsible for the temporal expression pattern of SPL genes is still unknown . To answer this question , we measured the abundance of the SPL2 , SPL3 , SPL9 , SPL10 , SPL11 , SPL13 and SPL15 transcripts in successive leaf primordia of the mir156a/c mir157a/c mutant ( Fig 8 ) . Most of these transcripts were present at either the same level or at slightly lower levels in adult LP ( LP5 , 6 , 9 , 10 ) compared to juvenile LP ( LP1 , 2 , 3 , 4 ) . The only exception was SPL3 , which increased 3–4 fold from LP1&2 to LP9&10 . This result suggests that miR156/miR157 are entirely responsible for the temporal increase in the SPL2 , SPL9 . SPL10 , SPL11 , SPL13 and SPL15 transcripts , whereas the temporal increase in SPL3 transcripts may be partly regulated by factors that operate independently of miR156/miR157 . Alternatively , the temporal increase in SPL3 may be attributable to the small amount of miR156/miR157 remaining in this quadruple mutant . The degree to which the abundance of different SPL transcripts changes in response to changes in the level of miR156/miR157 does not necessarily reflect the developmental importance of these SPL genes because miR156/miR157 can mediate both transcript cleavage [3 , 38–40] and translational repression [23–26] . For example , SPL3 is highly expressed in vegetative shoots and is more sensitive to miR156/miR157 than any other SPL gene , but the phenotype of spl3 mutations demonstrate that it plays little or no role in vegetative development [21] . We were particularly interested in determining whether SPL9 and SPL13 contribute to the precocious phenotype of mir156/mir157 mutants because SPL9 transcripts increase as miR156/miR157 levels decline , whereas SPL13 transcripts are relatively insensitive to changes in these miRNAs ( Fig 7 ) . To address this question , we introduced spl9 into a mir156a/c mutant background and introduced spl13 into a mir156a/c mir157a/c mutant background . spl9 completely suppressed the precocious abaxial trichome phenotype and partially suppressed the leaf shape phenotype of mir156a/c , whereas spl13 partially suppressed the effect of mir156a/c mir157a/c on both of these traits ( Fig 9 ) . Thus , SPL9 and SPL13 both play important roles in miR156-mediated developmental transitions . We studied the mechanism by which miR156/miR157 regulate the expression of SPL9 and SPL13 by comparing the abundance of the SPL9 and SPL13 mRNAs with the abundance of their protein products . Antibodies against SPL9 and SPL13 are not available , so we used previously described [21] and newly generated SPL9-GUS and SPL13-GUS translational reporters to visualize these proteins in transgenic plants . Leaf primordia were harvested sequentially as the shoot developed , and the abundance of the SPL9-GUS , SPL13-GUS and miR156 transcripts was measured by RT-qPCR , while the abundance of the SPL9-GUS and SPL13-GUS proteins was measured using the MUG assay . Consistent with previous results [21] , a nearly10-fold decrease in the level of miR156 between LP1&2 and LP9&10 was accompanied by very modest ( 2-fold or less ) increase in the level of the SPL9-GUS and SPL13-GUS transcripts ( Fig 10A and 10B ) . In contrast , the activity of the SPL9-GUS protein increased 10-fold between LP1&2 and LP9&10 ( Fig 10A ) , whereas the activity of SPL13-GUS protein increased 15-fold between LP1&2 and LP7&8 ( Fig 10B ) . The relationship between the change in miR156 levels and the change in SPL9-GUS and SPL13-GUS expression varied from leaf to leaf . The 4-fold decrease in miR156 between LP1&2 and LP3&4 was associated with a 3-fold increase in SPL9-GUS activity and a 9-fold increase in SPL13-GUS activity , but subsequent smaller changes in miR156 were associated with disproportionately large increases in the expression of these reporters . For example , in the SPL9-GUS line , miR156 declined by about 2-fold between P3&4 and LP9&10 , while the amount of SPL9-GUS protein increased 9-fold . In the SPL13-GUS line , miR156 declined by only 10% between LP3&4 and LP7&8 , while the amount of SPL13-GUS protein doubled . To examine the quantitative relationship between miR156 and SPL13 expression in more detail , we took advantage of a transgenic line containing an estrogen-inducible miR156 target-site mimic ( Ind-MIM156 ) , which enabled us to decrease the activity of miR156 by exogenous application of β-estradiol . One-week-old plants homozygous for the SPL13-GUS and In-MIM156 transgenes were given mock and β-estradiol treatments , and LP1&2 were harvested 24 hours later and analyzed by RT-qPCR and the MUG assay . This treatment reduced the abundance of miR156 by about 3-fold and produced a 2-fold increase in SPL13-GUS mRNA , but increased the abundance of the SPL13-GUS protein by greater than 15-fold ( Fig 10C ) . Because the amount of active miR156 in In-MIM156 may not be measured accurately by RT-qPCR , we also examined the abundance of SPL3 transcripts in mock- and estradiol-treated plants . The abundance of SPL3 mRNA is hypersensitive to variation in miR156 and thus serves as a proxy for the abundance of miR156 ( Fig 7 ) . SPL3 transcripts were 4-fold more abundant in induced plants relative to mock-treated plants ( Fig 10C ) , which is similar to difference in the amount of SPL3 transcripts in Col vs . mir156c ( Fig 7 ) . mir156c reduces miR156 by about 50% ( Fig 1B ) . Consequently , this result implies that estradiol-treated plants had approximately 50% less active miR156 than mock-treated plants , which is consistent with amount of miR156 detected by RT-qPCR . In summary , these results provide further evidence that miR156/miR157 regulate the expression of SPL13 primarily by promoting its translational repression , and also demonstrate that SPL13 activity responds non-linearly to changes in the abundance of these miRNAs . SPL9 transcripts are more sensitive to changes in miR156/miR157 than SPL13 transcripts ( Fig 7 ) , suggesting that transcript cleavage may play a larger role in the regulation of SPL9 than SPL13 . To address this possibility , we introduced the miR156-sensitive SPL9-GUS reporter into mir156a/c , mir157a/c , and mir156/c mir157a/c mutant backgrounds , and measured the abundance of the SPL9-GUS mRNA and protein in LP1&2 . mir157a/c did not have a significant effect on SPL9-GUS mRNA or protein levels , but mir156a/c produced a 2-fold increase in the SPL9-GUS transcript and a 5-fold increase in the SPL9-GUS protein ( Fig 10D ) . mir156a/c mir157a/c had an even more dramatic effect on the expression of SPL9-GUS , producing a 4-fold increase in the SPL9-GUS transcript and an ~36 fold increase in the SPL9-GUS protein ( Fig 10D ) . These results demonstrate that miR156/miR157 repress SPL9 both by destabilizing the SPL9 transcript and by repressing its translation . The increase in SPL9 activity that occurs during shoot development [21] is probably attributable primarily to a reduction in miR156/miR157-mediated translational repression because the SPL9-GUS protein increases more significantly in response to a decrease in miR156/miR157 than the SPL9-GUS transcript . To compare the sensitivity of the SPL9 and SPL13 transcripts to miR156/miR157-mediated cleavage , we used a modified form of 5’ RNA Ligase Mediated Rapid Amplification of cDNA Ends ( 5’ RLM-RACE ) [3 , 41] to quantify the ratio of un-cleaved/cleaved SPL9 and SPL13 transcripts in wild-type Col and mutants deficient for miR156 and miR157 . Equal amounts of total RNA from LP1&2 were ligated to a 5’-end RNA adaptor , and the purified RNA ligation products were then used in RT reactions using a poly-T primer . The levels of un-cleaved and cleaved SPL transcripts were then measured by qPCR , using primers specific for each type of transcript . These results were normalized to elf4A1 , and the un-cleaved/cleaved transcript ratio in each genotype was then calculated by dividing the relative expression values . This ratio does not necessarily reflect the actual difference between these transcripts because primers for un-cleaved and cleaved transcripts may have different amplification efficiencies . Consequently , instead of using this ratio to compare the relative abundance of cleaved SPL9 and SPL13 transcripts , we asked whether the cleavage of these transcripts is differentially sensitive to variation in the level of miR156/miR157 . This was done by normalizing the un-cleaved/cleaved transcript ratio from different mutants to the value in Col . The ratio of un-cleaved:cleaved SPL13 transcripts was about 2-fold greater in mir156a/c and mir156a/c mir157a/c than in Col , whereas the ratio of un-cleaved:cleaved SPL9 transcripts was 5-fold greater in mir156a/c and 15-fold greater in mir156a/c mir157a/c than in Col ( Fig 10E ) . Thus , SPL9 is more sensitive than SPL13 to miR156/miR157-directed transcript cleavage . The stoichiometry of a miRNA and its target can influence the mechanism of gene silencing [42] . To determine if the mode of action of miR156 is related to the relative abundance of miR156 and its targets , we measured the absolute quantity of several SPL transcripts and miR156 in LP3&4—the leaves in which the translational reporters for SPL3 , SPL9 , and SPL13 are first expressed [21] . This was done using known concentrations of SPL transcripts and miR156 as standards , and performing RT-qPCR on these standards in parallel with RNA from LP3&4 . There was a 5-fold range in the abundance of different SPL transcripts , with SPL5 and SPL15 being the least abundant , and SPL3 and SPL13 being the most abundant ( Fig 11 ) . miR156 was 100 times more abundant than SPL3 and SPL13 , about 200 times more abundant than SPL6 and SPL9 , and about 500 times more abundant than SPL5 and SPL15 ( Fig 11 ) . This result therefore suggests that greater than a 100-to-200-fold excess of miR156 is required to completely repress SPL genes . Assuming that the transcription rate of these SPL genes is the same in LP1&2 and LP3&4 , we predict that the amount of miR156 in LP1&2 ( where all miR156-regulated genes are completely repressed [21] ) is approximately 300–600 times greater than the amount of SPL3 and SPL13 transcripts , and approximately 1 , 500 times greater than the amount of SPL5 and SPL15 transcripts . Although the relative abundance of miR156 vs . SPL9 and SPL13 might suggest that translational repression is favored by a relatively low miR156:SPL transcript ratio ( SPL13 ) whereas transcriptional cleavage is favored by a high miR156:SPL transcript ratio ( SPL9 ) , this seems unlikely because a 90% reduction in the level of miR156 in mir156a/c produced only a slight increase in the level most SPL transcripts , including SPL9 ( Fig 6 ) . Indeed , we only observed a major increase in SPL transcripts in the mir156a/c mir157a/c quadruple mutant , implying that transcript cleavage does not require high levels of these miRNAs . Thus , the miR156/SPL transcript ratio cannot explain the difference in the sensitivity of the SPL9 and SPL13 transcripts to miR156-directed translational repression .
The vegetative period of shoot development is typically divided into two phases—a juvenile phase and an adult phase . However , many species display considerable morphological variation during these phases . In some species the first few leaves are referred to as "seedling leaves" because they are anatomically or morphologically distinct from other juvenile leaves [43–47] . In Arabidopsis , leaves 1&2 differ from other juvenile leaves in that they are smaller and rounder , have a less complex vascular system , and are less sensitive to exogenously applied gibberellin than other juvenile leaves [1 , 2 , 6] . Leaves 1&2 also have much lower levels of SPL proteins than other juvenile leaves [21] . We found that leaves 1&2 have a significantly more miR156/miR157 than other juvenile leaves , and that the largest absolute as well as relative decrease in these miRNAs occurs between leaves 1&2 and leaves 3&4 . Additionally , the amount of miR156/miR157 in leaves 1&2 far exceeds the amount that is actually required to determine their identity; mutations that nearly completely eliminate miR156 have a similar effect on leaves 3 and 5 , but a much weaker effect on leaf 1 . The only genotypes that caused leaf 1 to resemble leaves 3 and 5 were those that reduce miR156/miR157 by 90% . These results suggest that leaves 1&2 represent a distinct developmental phase . Starting with leaf 3 , leaf size , the number of leaf serrations , and the angle of the leaf base change gradually from leaf-to-leaf . These gradual changes in leaf morphology are accompanied by an increased ability to produce abaxial trichomes [1] , which first appear between leaf 6 to 9 . In contrast to the morphological stability of leaves 1&2 , the morphology of these “late” juvenile leaves is influenced by light intensity , photoperiod , and the reproductive state of the shoot [2 , 5] . This combination of quantitative and qualitative changes , as well as the morphological plasticity of late juvenile leaves , can be explained by the relatively low and gradually decreasing level of miR156/miR157 in successive leaves , and by the non-linear response of some SPL genes to changes in the abundance these miRNAs . Features of leaf morphology that change continuously from leaf-to-leaf are likely to be controlled by pathways or processes whose activity is directly correlated with the level of SPL gene expression , whereas all-or-none traits , such trichome initiation , may only appear when the expression of these genes exceeds a threshold . Our results also explain why the expression of phase-specific traits becomes dissociated under certain conditions . For example , abaxial trichome production is more responsive to conditions that promote floral induction than either hydathode number or leaf shape [2] . Juvenile and adult vegetative traits can also be dissociated in English ivy , resulting in plants that display different combinations of these traits [48 , 49] . We suspect that this phenomenon is attributable to functional differentiation between SPL genes , coupled with variation in their sensitivity to miR156 and miR157 . Some SPL genes are expressed at relatively high levels and respond nearly linearly to changes in the level of miR156/miR157 , whereas others are expressed at relatively low levels and only respond significantly to a change in the level of miR156/miR157 when these miRNAs are present at very low levels . Given that SPL genes are not functionally identical [9 , 21 , 50] , conditions that produce small changes in miR156/miR157 , or which elevate the transcription of particular SPL genes above the threshold established by miR156/miR157 , could lead to unusual combinations of phase-specific traits . miR156 is one of the oldest and most highly conserved miRNAs in plants [51 , 52] . Although miR157 is nearly as old and as highly conserved as miR156 , this is not widely appreciated because miR157 is frequently annotated as miR156 in small RNA sequencing studies and in miRBase ( http://www . mirbase . org ) . The failure to distinguish these miRNAs is likely based on the assumption that they have the same function . Our comparison of the expression patterns of miR156 and miR157 , as well the phenotype of plants lacking one or both of these miRNAs , demonstrates that these miRNAs work together to regulate vegetative phase change , but are not functionally identical . miR157 is more abundant than miR156 and is expressed in a similar temporal pattern , but miR156 plays a more important role in vegetative phase change and is a more potent repressor of SPL gene expression . The difference in the activity of these miRNAs may be due , in part , to the lower efficiency with which miR157 is loaded onto AGO1 . However , this is not the only reason for the difference in their activity because the amount of miR157 associated with AGO1 is not dramatically lower than the amount of miR156 associated with AGO1 . Another factor that may contribute to the difference in their activity is the structure of the miRNA:target-site duplex [41 , 53–55] . miR156 and miR157 bind to most of their targets with a single mismatch , but this mismatch is located one nucleotide from the cleavage site in the case of miR157 and 3 nucleotides from the cleavage site in the case of miR156 . miR157 also has an additional 5' nucleotide ( relative to miR156 ) , which is unpaired in the miR157:SPL13 duplex . SPL13 plays a major role in vegetative phase change [21] and if this mismatch reduces the ability of miR157 to repress the activity of SPL13 , this would be expected to have a significant phenotypic effect . We suspect that this extra 5' uracil is primarily responsible for the relatively low activity of miR157 because miR156d also has an extra 5' uracil and is significantly less active than miR156 , despite being otherwise identical to miR156 . However , we cannot rule out the possibility that the difference in the activity of miR156 and miR157 is a consequence of the difference in their length , rather than the specific features of the miRNA:SPL duplex . Most miRNAs in plants are 21 or 22 nt in length , but several evolutionarily conserved miRNAs are 20 nt , or exist as both 20 nt and 21 nt variants [52 , 56] . miRNAs that are 22 nt are uniquely capable of generating tasiRNAs and other types of phased siRNAs [57 , 58] , but it is still unknown if 20 and 21 nt miRNAs are functionally distinct . The 20 nt miR156 transcript is present in the moss , Physcomitrella patens [59 , 60] , and in virtually all other plants that have been examined to date [51 , 52] . miR157 is absent in Physcomitrella , but is present in Selaginella and most , but not all , higher plants [61] . The fact that miR157 has been conserved along with miR156 during plant evolution suggests that it is not completely redundant with miR156 , and raises the possibility that the relative activity of these miRNAs may differ in different species . The effect of mutations in different MIR156 and MIR157 genes on the abundance of miR156 and miR157 demonstrates that MIR156A and MIR156C produce most of the miR156 in the shoot whereas MIR157C produces most of the miR157 . MIR156D and MIR157A are expressed at a much lower level than these loci , but the ability of mir156d and mir157a to enhance the phenotype of plants mutant for mir156a , mir156c , and mir157c demonstrates that they are functionally significant . We do not know if the miR156 and miR157 transcripts that remain in the mir156a/c/d mir156a/c mutant are derived from one or more these loci or from other MIR156/MIR157 genes because we cannot be certain that the mutations present in this mutant stock are completely null . Whatever the case , these remaining transcripts are functionally active because the phenotype of this pentuple mutant is not as strong as the phenotype of plants over-expressing a miR156 target site mimic [62] . A complete picture of the function of this gene family will require identifying loss-of-function mutations in MIR156E , F , G and H , and MIR157B and D . The phenotype of slicer-defective AGO1 mutants suggests than plant miRNAs destabilize transcripts exclusively by transcript cleavage , and that this is the primary mode by which they regulate gene expression [63 , 64] . However , the results of this and previous studies [23–26] indicate that miR156 and several other plant miRNAs act primarily by promoting translational repression [24–27 , 65 , 66] . This observation begs the question of why translational repression is so important for the function of these miRNAs , and how the choice between transcript cleavage and translational repression is regulated . Our results suggest that the amount of miR156 and miR157 present in both juvenile and adult leaves is sufficient to almost completely saturate the cleavage machinery at most of their targets . However , the response of different SPL transcripts to miR156/miR157 varies between transcripts , suggesting that the susceptibility of these transcripts to miR156/miR157-induced cleavage depends on sequences outside the miR156/miR157 target site . In plants , the importance of sequences flanking a miRNA target site has been demonstrated for miR159 [42 , 67] and for several miRNA-cleaved transcripts that generate phasiRNAs [68] . However , it is unclear if the sequence environment of a miRNA target site influences the mechanism by which a miRNA represses gene expression . A comparison of the molecular mechanism by which miR156/miR157 regulate the expression of SPL9 and SPL13 will be informative because these genes respond very differently to changes in the level of these miRNAs . miR156/miR157 are among the oldest miRNAs in plants , and it is therefore reasonable to conclude that miRNA-induced translational repression is an ancient regulatory mechanism in plants . Identifying the biochemical factors that induce AGO1 to direct transcript cleavage vs . translational repression , and defining the functional consequences of these modes of regulation , are important problems for future research .
All of the lines used in this study were in a Col genetic background . The mir156a-2 and mir156c-1 mutations have been described previously [69] . mir156d-1 ( SALK_40772 ) , mir157a-1 ( Flag_375C03 ) , mir157c-1 ( SALK_039809 ) were obtained from the Arabidopsis Biological Resource Center ( Ohio State University , Columbus , OH ) and were crossed to Col at least 3 times before further analysis . mir156b-1 was generated by TALEN-directed mutagenesis [70] in a mir156c-1 background , and is a 42 nt-deletion within the MIR156B hairpin sequence ( AACAGAGAAAACTGACAGAA—-42 bp deletion—GCGTGTGCGTGCTCACCTCTC ) that removes most of the miR156 sequence . Multiple mutant lines were generated by inter-crossing mutations and then screening F2 populations for the desired genotypes using the allele-specific primers listed in S1 Table . Seeds were sown on Farfard #2 Mix and placed at 4° C for 3 days before moving to a Conviron growth chamber , where they were grown under either long day ( 16 hrs light/8 hrs dark; 80 μmol m-2 s-1 ) or short day ( 10 hrs light/ 14 hrs dark; 130 μmol m-2 s-1 ) conditions , with illumination provided by a 6:2 ratio of broad spectrum ( Interlectric Tru-lite ) and red light-enriched ( Interelectric Gro-lite ) fluorescent lights . The miR156-sensitive and miR156-resistant SPL13-GUS reporter lines used in this study were described previously [21] . The previously described SPL9-GUS reporter lines [21] silenced when they were crossed into a miR156a-2 background , so it was necessary to produce new lines for these reporters . For this purpose , miR156-sensitive and miR156-resistant SPL9:SPL9-GUS genomic sequences [21] were inserted into the pCAM-NAP:eGPF vector [71] using the restriction enzymes XmaI and SbfI . These constructs were then introduced into the miR156a-2/miR156c-1 miR157a-1/miR157c-1 lines by Agrobacterium-mediated transformation . Homozygous single insertion lines were selected as described previously [71] , and crossed to Col and further genotyped to obtain SPL9-GUS reporters in different genetic backgrounds . The estradiol-inducible MIM156 line ( Ind-MIM156 ) was constructed using a Gateway compatible version of the XVE system , as described by Brand and colleagues [72] . The MIM156 sequence described by Franco-Zorilla and colleagues [53] was cloned into pMDC160 by standard Gateway cloning using the primers in S1 Table ( referred to as pMDC160-MIM156 ) . Plants containing pMDC150-35S [72] were crossed to transgenic pMDC160-MIM156 plants and made homozygous . Induction of gene expression was performed by spraying 10μM 17-ß-estradiol ( 0 . 01% Silwet 77 ) on seedlings at the desired time point . Tissues were harvested at 24hr after induction . Tissue samples were harvested into 2ml tubes submerged in liquid nitrogen , and then homogenized using a bead-beater . 300μl of extraction buffer ( 10 mM EDTA pH 8 . 0 , 0 . 1% SDS , 50 mM sodium phosphate pH 7 . 0 , 0 . 1% Triton X-100; 10 mM ß-mercaptoethanol and 25 μg/ml PMSF added fresh before experiment ) was then added to each tube . Samples were mixed well and incubated on ice for 10 mins , after which they were centrifuged at 4°C ( 13000 rpm ) for 15 mins . to remove cell debris . 96ul of supernatant was removed and incubated with 4ul of 25mM 4-MUG at 37°C . Incubation time varied among reporters to ensure the end fluorescence readings fell within a linear range . The reaction was terminated by adding 100ul of 1M sodium carbonate to each tube , and fluorescence was measured using a Modulus fluorometer ( E6072 filter kit ) . The amount of MU in each sample was then calculated by comparing this reading to a standard curve constructed by plotting the fluorescence readings from serial dilutions ( 100nM , 250nM , 500nM , 1000n ) of 4-MU . The 4-MU equivalent was divided by the incubation time and this value was then normalized to the amount of protein in the sample , which was determined by performing a Bradford assay on the supernatant remaining in the original tube . For each sample , GUS activity was expressed as 4-MU equivalent/min/mg protein . Values were then normalized to the control sample of each experiment . RNA was extracted from leaf primordia no larger than 1mm in length using Trizol ( Invitrogen ) , and samples were then treated with DNase ( Ambion ) following the manufacturer’s instructions . To measure the abundance of miRNAs , 600ng of RNA was used in a reverse transcription reaction with a SnoR101 reverse primer and a miRNA-specific RT primer . To measure the abundance of SPL transcripts , 600ng RNA was used in a reverse transcription reaction primed with Oligo ( dT ) . qPCR was performed on the resulting products , using the primers listed in S1 Table . Reactions were performed in triplicate for each biological replicate . Tissue was homogenized in liquid nitrogen , and total RNA was then extracted using Trizol ( Invitrogen ) . Extracts were incubated in 500mM NaCl and 5% PEG8000 on ice for 2 hours , and centrifuged at 13 , 000 rpm for 10min . The supernatant was incubated with a 10% volume of 3M NaOAc and 2 volumes of 100% ethanol at -20°C for 2 hours . Small RNAs were precipitated by centrifugation at 13 , 000 rpm for 10min , and washed in cold 75% ethanol twice . RNA blotting was performed as described previously [3] . A 1:1 ratio of miR156 and miR157 probes was used for mixed probe hybridizations . Sequencing libraries were generated from small RNAs isolated from shoot apices of FRI FLC and FRI flc-3 seedlings grown in the conditions described by Willmann and colleagues [2] . The shoot apex samples consisted of the shoot apical meristem and leaf primordia 1 mm or less in length . Libraries were generated using a lab-assembled version of Illumina's 2007 small RNA library sample preparation protocol , followed by high-throughput sequencing with Illumina's Genome Analyzer II platform . The miR156 and miR157 transcripts used as references were synthesized by IDT , and the SPL transcripts used as references were synthesized by in vitro transcription . The template for each in vitro transcription reaction was generated by PCR , using the primers listed in S1 Table and cDNA from Col . Each purified SPL transcript was assayed by denaturing gel electrophoresis to confirm that the in vitro transcription product was a single species of the expected size . To quantify SPL transcripts , the reference mRNA generated by in vitro transcription was diluted to 1 . 00E-8 M and this sample was then used to create a 10x dilution series in 600ng/μl total RNA from E . coli . This dilution series was analyzed by RT-qPCR in parallel with RNA isolated from LP3&4 . A series of 2x dilutions of the reference mRNA sample whose concentration was similar to that of the experimental sample was then constructed , and run along with the experimental sample in a second RT-qPCR reaction . The 2-CT values of the reference samples were plotted against their known concentrations , and the CT value of the unknown sample was then placed on this graph to determine the RNA concentration . Ligation reactions were performed with 5 μg of total plant RNA and 1 μg of the GeneRacer ( Invitrogen ) RNA adapter following the manufacturer’s instructions , but without carrying out the de-capping reaction . After 2 hrs of incubation at 37°C , the reaction mixture was diluted with nuclease- free water and RNA was extracted in phenol: chloroform . The purified ligation product was dissolved in 10μl nuclease-free water , and 5μl of this solution was used in a reverse transcription reaction with an oligo ( dT ) primer . qPCR was performed using primers listed in S1 Table to quantify cleaved and un-cleaved SPL transcripts . Two-week-old seedlings were harvested in liquid nitrogen and homogenized in a cold motar and pestle . For each sample , approximately 1mL ground powder was dissolved in 2mL lysis buffer ( 50mM Tris HCl , pH 7 . 4 , with 150mM NaCl , 1mM EDTA , 1% Triton X-100 , 1mM PMSF , 1% Protease Inhibitor ) followed by 15 min incubation on ice . 20% of the homogenized sample was saved for RNA extraction , and the rest was centrifuged at 13 , 000 rpm at 4°C for 20min to remove cell debris . The resulting supernatant was then filtered through a 45μm filter . Immunoprecipitation was performed using Anti-FLAG M2 Magnetic Beads ( Sigma ) following the manufacturer’s instructions . RNA was extracted from the beads using Trizol ( Invitrogen ) and analyzed by Northern blotting . Small RNA sequence data are available in the NCBI Gene Expression Omnibus database under series accession number GSE72303 . | Leaves produced at different stages in the development of an Arabidopsis shoot vary predictably in shape and size . Previous studies have shown that this phenomenon is regulated by variation in the abundance of the miRNAs , miR156 and miR157 , but how miR156/miR157 produce the changes in leaf morphology that occur during shoot development is not understood . To answer this question , we measured the abundance of miR156/miR157 and their SPL targets in successive leaf primordia , and characterized the effect of variation in the abundance of miR156/miR157 on leaf morphology and the abundance of SPL transcripts and SPL proteins . miR156/miR157 are present at very high levels in the first two rosette leaves , where they act as buffers to stabilize leaf identity . They are present at lower and steadily declining levels in subsequent leaves , where they act to modulate leaf morphogenesis . In these later-formed leaves , a small decrease in the abundance of miR156/miR157 produces a disproportionately large increase in SPL activity , primarily as a result of the increased translation of SPL transcripts . Our results provide a new view of vegetative phase change in Arabidopsis and the mechanism by which miR156 and miR157 regulate this process . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"plant",
"anatomy",
"trichomes",
"molecular",
"probe",
"techniques",
"gene",
"regulation",
"plant",
"science",
"micrornas",
"molecular",
"biology",
"techniques",
"seedlings",
"rna",
"sequencing",
"plants",
"research",
"and",
"analysis",
"methods",
"probe",
"hybridization",
"gene",
"expression",
"leaves",
"molecular",
"biology",
"biochemistry",
"rna",
"eukaryota",
"nucleic",
"acids",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"organisms"
] | 2018 | Threshold-dependent repression of SPL gene expression by miR156/miR157 controls vegetative phase change in Arabidopsis thaliana |
Adaptive evolution in humans has rarely been characterized for its whole set of components , i . e . selective pressure , adaptive phenotype , beneficial alleles and realized fitness differential . We combined approaches for detecting polygenic adaptations and for mapping the genetic bases of physiological and fertility phenotypes in approximately 1000 indigenous ethnically Tibetan women from Nepal , adapted to high altitude . The results of genome-wide association analyses and tests for polygenic adaptations showed evidence of positive selection for alleles associated with more pregnancies and live births and evidence of negative selection for those associated with higher offspring mortality . Lower hemoglobin level did not show clear evidence for polygenic adaptation , despite its strong association with an EPAS1 haplotype carrying selective sweep signals .
Understanding the impact of natural selection on phenotypic variation has been a central focus of evolutionary biology since its beginning as a modern scientific discipline . Decades of research have accumulated evidence for widespread adaptive phenotypic evolution in nature , including correlations between phenotypes and environmental factors [1–3] , and higher reproductive success of native individuals compared to visitors [4] . Beyond the phenotypic studies , much effort has been devoted , especially in humans , to identifying adaptive alleles through indirect statistical approaches that use genetic variation data and that can detect the impact of past selective pressures [5] . The most widely used family of approaches aims at detecting new beneficial mutations that were quickly driven to high frequency or fixation by natural selection , a model that is often referred to as selective sweep and that is likely to apply to adaptive alleles of large effect [5–9] . However , genome-wide association studies have revealed that most phenotypic variation in humans is highly polygenic; in other words , it is due to the combined effects of a large number of alleles with small effects [10–12] . Under this scenario , adaptations will tend to generate upward shifts in the frequency of adaptive alleles at many loci rather than a major shift at one or few loci , as is the case , for example , for lactase persistence . Methods for detecting polygenic adaptations combine two sources of information: genome-wide association studies ( GWAS ) provide alleles associated with a phenotype of interest as well as their effect size , and the population frequency of GWAS alleles enable inter-population comparison [13–15] . These indirect methods can provide evidence for past selective events , but each is sensitive to different selection models [16 , 17] , thus providing insights into a subset of adaptive alleles [18] . Moreover , these approaches cannot distinguish among selective effects on different fitness components , e . g . fertility vs . viability . A major advantage of indirect approaches is that they can detect selective sweep signals due to plausible , low selection coefficients ( as long as 4Nes > 1 ) with comparatively small sample sizes . A complementary set of approaches aims at assessing directly the effects of genotype on reproductive fitness [19] . These direct approaches have many advantages , mainly the ability to detect selective events occurring in the present generation and the similar sensitivity to different selection models , e . g . balancing vs . directional selection [20] . However , they require large sample sizes to detect plausible selective coefficients . Large cohorts with genetic information are becoming increasingly available for humans , enabling approaches that were not feasible until recently [21 , 22] . For example , a recent study analyzed two cohorts , for a total of more than 175 , 000 individuals , to assess genetic effects on viability by identifying alleles that consistently changed in frequency with age [23] . Another direct approach is to search for variants influencing fitness through GWAS of reproductive traits such as number of children ever born [24] , twinning rate or mother’s age at first birth [25] . However , the genetic bases of reproductive traits remain markedly understudied , despite their great evolutionary and biomedical significance [26] . High altitude populations have emerged as an ideal system to study the genetic architecture of human adaptations . Populations of the high-altitude regions of Tibetan , Andean , and East African Plateaus have been exposed to the stress of hypobaric hypoxia for sufficient time [27] to have allowed the evolution of new adaptive traits [28–32] . Recent population genomic studies of Tibetans detected strong selective sweep signals in Tibetans at two loci , EGLN1 ( egl-9 family hypoxia inducible factor 1 ) and EPAS1 ( endothelial PAS domain containing protein 1 ) [33–35] , each coding for a key component of the regulatory program responding to variation in oxygen supply [36] . Importantly , alleles in these genes that occur at high frequency in Tibetans but are rare elsewhere were also reported to be associated with lower hemoglobin concentration ( Hb; g/dL ) [33–35 , 37] ( but see [37–39] ) , consistent with many observations that unelevated Hb is characteristic of high-altitude Tibetans ( reviewed in [32] ) . Because the impact of hypobaric hypoxia on human physiology cannot be modified through behavioral or cultural practices , indigenous high altitude populations provide a rare opportunity to observe human evolution in action . Here , we took advantage of this property to design a study aimed at comprehesively dissecting adaptations to high altitude using both direct and indirect approaches . Our goal was to map , in the same sample , physiological as well as reproductive variables and to apply polygenic adaptation tests using information about the alleles associated with these traits . To this end , we collected genotype and phenotype data in a sample of ethnically Tibetan women in post-reproductive age ( so that they had completed their family size ) . We found a single genome-wide significant association signal for oxygenated hemoglobin ( “oxyHb” ) at the EPAS1 locus and several signals for reproductive traits . We tested for selective events that took place in the past , through indirect approaches that can detect polygenic adaptations , as well as for ongoing events , through the direct approach of mapping measures of reproductive success . We detected signatures of polygenic adaptation for reproductive traits such as numbers of livebirths and offspring mortality , consistent with selective processes that are still ongoing in contemporary populations . In contrast , we found little evidence for polygenic adaptations toward lower Hb .
In the context of studies of high-altitude adaptation , the term “Tibetan” refers to the modern descendants of the ancient indigenous population of the Tibetan plateau who share cultural and biological affinities and reside in several polities , including Nepal . To investigate the genetic bases of high-altitude adaptations in Tibetan populations , we collected physiological and reproductive phenotype data and saliva samples of 1 , 008 indigenous ethnically Tibetan women living at 3 , 000–4 , 000 m in the Mustang and Gorhka districts of Nepal ( see Materials & Methods ) . All Tibetan participants were chosen to be 39 years of age or older so that their recorded reproductive history would have minimal confounding due to unrealized reproduction . We also obtained saliva samples for DNA extraction and analysis from 103 Sherpa participants ( including ten parents and offspring trios ) from the high-altitude regions in the Khumbu district in Nepal . The Sherpa data were included in the reference panel for genotype imputations and in the polygenic adaptation tests , but not in GWAS ( see below ) . Genetic variation data of our study participants were generated by a combination of experimental and computational tools ( Fig 1; also see Materials & Methods ) . First , we generated novel genotype data for all participants using Illumina genotyping array platforms in multiple phases ( S1 Table ) . Briefly , all Tibetans were first genotyped for about 300K markers on the HumanCore array with additional 2 , 553 custom markers to cover candidate regions , including the EGLN1 , EPAS1 , HIF1A ( hypoxia inducible factor 1 alpha subunit ) and NOS2 ( nitric oxide synthase 2 ) genes . Then , we genotyped a subset of 344 unrelated Tibetans ( allowing up to first cousins ) and all 103 Sherpa for over 700K markers on the OmniExpress array; the same individuals were separately genotyped for two nonsynonymous SNPs in the EGLN1 gene , rs12097901 and rs186996510 [40] . Analyzing all Tibetan individuals together with published genome-wide genotype data of world-wide ancient and modern populations , we show that these Nepali Tibetans are genetically most closely related to other high altitude East Asians , such as Sherpas [41] and Tibetans from Lhasa [42] , that they can be modeled as a mixture of Sherpas and South and Central Asians ( e . g . Pathans ) , and that they derive on average only 3 . 4% of their ancestry from the latter gene pool ( ranging 0 . 0–11 . 6%; S1 Text ) . To augment publicly available reference panels for genotype imputation , we generated whole genome sequence data of 18 Sherpa and 35 Tibetans ( S1 Table ) . Three Sherpa trios and four Tibetan mother-daughter duos were sequenced to high coverage ( ~ 20x ) , while the remaining 36 individuals were genetically unrelated and sequenced to low coverage ( ~ 5x , S1 Table ) . For sequencing , we chose Tibetan and Sherpa individuals with no signature of recent admixture ( See Materials & Methods ) . Adding six previously published Tibetan and Sherpa genomes [41 , 43] , we obtained phased genotypes of 59 individuals including 9 , 742 , 498 variants , of which 1 , 364 , 150 were not found in the 1000 Genomes Project ( 1KGP ) phase 3 data set [44] . Among the non-1KGP variants , 540 , 218 were included in dbSNP150 database , while 823 , 932 were not . Variant annotation using the ANNOVAR program [45] identified 8 , 679 nonsynonymous variants , 235 nonsense coding variants , and 126 splicing variants not present in the 1KGP ( S2 Table ) . Among the non-1KGP variants , 29 . 46% and 24 . 14% occurred as singletons and doubletons , but 11 . 06% of them occurred at frequency 10% or higher ( S2 Table ) . Using both the 1KGP phase 3 data and our high altitude sequence data as reference panels , we performed genotype imputation of all samples using the IMPUTE2 program [46] to generate the analysis-ready genotype data of 991 individuals covering about 3 . 5 million SNPs in Tibetans ( see Materials & Methods ) . We performed GWAS of 23 phenotypes characterizing the reproductive history of our study participants using a linear mixed model-based approach as implemented in GEMMA [47] . Although they are partially correlated , these traits provide a comprehensive assessment of reproductive fitness and , importantly , allow evaluating the effects of selection on viability and fertility components , separately . We grouped our fertility phenotypes into two categories , “fertility counts” ( e . g . number of live births ) and “fertility proportions” ( e . g . proportion of live births among pregnancies ) ( Table 1 and S3 Table describe the sample and summarizes the reproductive phenotypes ) . While the count phenotypes are more directly related to evolutionary fitness , they may be confounded by compensatory reproduction in case of a negative pregnancy outcome or by sociocultural factors that influence the count [48] . In contrast , the proportional variables are less affected by such factors and may provide information on the specific phase of the reproductive process affected by the associated genetic variation . Therefore , the counts and proportions may capture different aspects of the reproductive outcome . The GWAS was performed on the entire sample and on a subset referred to as continuously married ( CM ) , that was composed of about 60% of participants who had been in a marital relationship throughout the ages of 25 to 40 ( see Materials & Methods ) . This subset controls the variance in marital relationship status; on the other hand , the resulting smaller sample size reduces the power to detect significant associations . For the fertility count phenotypes , we found 55 SNPs with genome-wide significant genotype-phenotype associations , which can be reduced to six association peaks when linkage disequilibrium among associated SNP is taken into account ( Table 2 ) . These six peaks reflect five independent association signals , considering that the number of pregnancies and the number of live births are highly correlated . First , our analysis of three fertility count phenotypes yielded genome-wide significant association peaks . Two intronic SNPs in the CCDC141 ( coiled-coil domain containing 141 ) gene , with the top SNP rs6711319 , were associated both with the number of pregnancies ( p = 2 . 10×10−8 ) and with the number of live births ( p = 2 . 89×10−9; Fig 2 and S1 Fig ) . Fourteen SNPs between the PAPOLA ( poly ( A ) polymerase alpha ) and the VRK1 ( vaccinia related kinase 1 ) genes were associated with the number of stillbirths in the continuously married subset ( p ≥ 8 . 38×10−9; Fig 2 and S1 Fig ) . The same set of SNPs also showed a suggestive association in the complete sample set ( “All” ) , including all individuals regardless of whether they were continuously married ( p = 6 . 23×10−5 to 1 . 85×10−4 ) . No expression quantitative trait loci ( eQTLs ) were detected in the GTEx Project data in this peak region [49] , making it hard to connect the associated SNPs with a specific gene . For the fertility proportion phenotypes , two genome-wide significant association signals were detected ( Table 2 ) . Eight SNPs near C6orf195 , with the top SNP rs9392394 , were associated with the proportion of children who died before age 15 ( p = 1 . 05×10−8 ) ; SNPs within 15 kb of this peak had been associated with heart , blood pressure and reticulocyte traits in GWAS [50–52] . Twenty-nine SNPs near CTBP2 , with the top SNP rs1459385 , were associated with the same phenotype in the continuously married subset ( p = 1 . 59×10−8; Fig 2 and S1 Fig ) ; other genes in this region include TEX36 ( testis expressed 36 ) and EDRF1 ( erythroid differentiation regulatory factor 1 , which regualtes the expression of globin genes ) . No eQTLs were detected in the GTEx Project data in these association regions . The genetic bases of Hb , percent of oxygen saturation of hemoglobin ( SaO2 ) , and pulse have been previously studied in outbred populations mainly of European ancestry [52–58] . Here , we performed GWAS of these key physiological phenotypes in Tibetans , measured by the same non-invasive device , to potentially uncover population-specific genetic determinants . We derived two additional composite phenotypes: oxygenated hemoglobin concentration ( “oxyHb” , defined as the product of Hb and SaO2 divided by 100 ) , and deoxygenated hemoglobin concentration ( “deoxyHb” , defined as the difference between Hb and oxyHb ) . Consistent with findings from other studies , these women had an average hemoglobin concentration of 13 . 8 g/dL ± 1 . 3 g/dL ( mean ± 1 standard deviation ) . Table 1 and S3 Table describe the sample and summarize the phenotypic data . Each GWAS included about 3 . 5 million SNPs with minor allele frequency ( maf ) ≥ 0 . 05 . Eight SNPs within a 17 kb intronic region of the EPAS1 gene were significantly associated with oxyHb ( p ≤ 5 × 10−8 for all eight SNPs , with the top signal at rs372272284; Table 2 , Fig 2 and S2 Fig ) . Hb and oxyHb were strongly correlated ( Pearson r = 0 . 874 ) , and all eight SNPs were also strongly associated with Hb ( p ≤ 4 . 10×10−7; S4 Table ) . This is the first report of an association of the derived Tibetan EPAS1 alleles with a hemoglobin trait that reaches genome-wide significance levels . The results , including the estimated effect size of 0 . 332 g/dL per allele , support previous candidate gene studies for Hb [33 , 34] . Due to strong linkage disequilibrium ( LD ) , the signature of a selective sweep around the EPAS1 gene in Tibetans extends farther than 100 kb; however , our large sample size and dense genetic variation data allowed us to narrow down the association signal to a 17 kb region . Conditioning on the genotype of rs372272284 , no residual association with either Hb or oxyHb was observed in the EPAS1 locus ( p ≥ 0 . 770; S3 Fig ) . This includes a previously identified “Tibetan-enriched” deletion ( “TED” ) , 81kb downstream of the EPAS1 gene , present in Tibetans but not in the introgressed Denisovan haplotype [59] . TED is in LD with the eight significant SNPs in our data set ( Pearson r = 0 . 771–0 . 783 ) , but its association with Hb and oxyHb was much weaker than our top SNPs ( p = 1 . 64×10−3 and 3 . 09×10−4 , respectively ) . In contrast , we did not detect significant associations in the EGLN1 gene , even when we added menopause status , a female-specific covariate of Hb , as an additional covariate or confined our analysis to post-menopausal women ( p ≥ 0 . 641 ) . Moreover , our data showed no significant interaction between the EPAS1 and EGLN1 SNPs ( rs372272284 and rs186996510 , respectively ) in the association with Hb ( marginal effect for the EGLN1 SNP rs186996510 , p = 0 . 196 , effect size β = 0 . 095 ± 0 . 074 g/dL; interaction effect , p = 0 . 613; β = 0 . 057 ± 0 . 114 ) [60] . Our results in post-reproductive females are consistent with and reinforce recent evidence suggesting that if EGLN1 SNPs affect Hb , they do so only in males ( S2 Text ) [37–39] . To see if the phenotype-associated markers are under strong positive selection , we used the population branch statistic ( PBS ) [34] with 1KGP phase 3 CHB ( Han Chinese in Beijing , China ) as a comparison group and 1KGP phase CEU ( CEPH Utah residents with Northern and Western European ancestry ) as an outgroup . Based on the allele frequency differentiation , PBS measures the level of allele frequency change specific to the target population ( i . e . Tibetan ) that is not shared by its comparison group . We find no selective sweep signal over the genome-wide significant association peaks except for EPAS1 , and find no loci with strong signatures of selective sweep beyond EPAS1 ( PBS = 1 . 073 , rs73926264 ) and EGLN1 ( PBS = 0 . 797 , rs186996510; S3 Text and S5 Table ) . A previous analysis of this sample of Tibetan women found strong relationships between physiological traits and reproductive success in this Tibetan sample by using a large set of covariates , including physiological , sociocultural , and socioeconomic variables ( e . g . relative wealth rank , type of marriage , and marital status ) [48] . Because the genetic analyses performed here used only a subset of those covariates , we tested for association of physiological traits and reproductive success by correcting for the specific set of covariates used in our GWAS . Consistent with the previous analysis , we found that lower Hb ( in females in post-reproductive years ) correlated with a higher proportion of live births among pregnancies ( p = 0 . 002 ) . We also found that Hb correlated positively with the numbers of stillbirths or miscarriages ( p = 0 . 040 and 0 . 057 , respectively ) , as well as their proportions among pregnancies ( p = 0 . 023 and 0 . 033 for stillbirths and miscarriages , respectively ) . Another interesting finding was the negative correlations between pulse and most of the fertility phenotypes , with the strongest correlations found with the numbers of pregnancies and livebirths ( p = 2 . 02×10−5 and 2 . 76×10−5 , respectively; S6 Table ) . Pulse’s negative association with a woman’s age at her last pregnancy partially accounts for this strong correlation; however , the association between pulse and the number of pregnancies remained significant ( p = 0 . 005 ) after correcting for a woman’s age at her last pregnancy , even if weaker . The pulse and fertility traits were previously shown to be marginally correlated if a larger set of physiological and sociocultural covariates was included in the model ( p = 0 . 130 and 0 . 069 for the numbers of pregnancies and live births , respectively ) [48] . These results suggest that both low Hb and low pulse are associated with better reproductive outcomes at high altitude and raise the possibility that genetic variants decreasing these traits were selected for in Tibetans . The women’s reproductive history data offer a unique opportunity to ask if selective sweep signals are associated with ongoing selection in contemporary Tibetans due to maternal factors . Consistent with previous studies , the EPAS1 and EGLN1 loci harbored the highest PBS values ( S5 Table ) . However , we did not detect an association between EGLN1 and EPAS1 SNPs and any of the direct measures of fitness; nominal levels of significance were observed in some cases , but no test reached significance after multiple test correction . Power calculations for the fertility count phenotypes suggest that we can detect such an association only if the associated selection coefficient is extremely high ( ≥ 6 . 6% per allele for 80% power given a single test; S7 Table ) and well above previous estimates for both EPAS1 and EGLN1 , 1 . 5% and 2 . 9% , respectively [37 , 61] . Therefore , these results do not rule out that these variants are advantageous . Other SNPs with high PBS values also did not show a significant association with fertility variables . To test for polygenic adaptations for low Hb and low pulse , we used two methods specifically designed to detect consistent changes in the frequency of alleles at many independent GWAS SNPs for a trait of interest . The first approach considers the frequency difference of the trait-increasing alleles between pairs of populations , specifically Tibetans or Sherpa and 1KGP CHB; the results are compared to 10 , 000 sets of control SNPs [13] . The second , more recent approach [14] calculates a genetic value for a trait of interest in each population by summing up the product of the frequency at each GWAS SNP and the effect size of that SNP and it compares GWAS-ascertained SNPs with a large number of control SNPs . Specifically , we focused on two tests . The “overdispersion” test asks if allele frequencies of the GWAS SNPs as a group show either unusually big differentiation across populations or unexpectedly strong correlation in the direction of change . The “outlier” test asks if the genetic value of a trait in a population or a group of populations is significantly different from that of the other populations . To test for positive selection for lower Hb , we used the 36 and 43 independent SNPs ( p ≤ 10−4 ) ascertained from our Tibetan Hb and oxyHb GWAS , respectively . Compared to control SNPs [13] , the Hb SNPs identified in Tibetans showed on average a lower frequency of Hb-increasing alleles in both Tibetans and Sherpa , suggesting selection favoring lower Hb levels ( one-sided empirical-p = 0 . 047 and 0 . 018 , respectively; Fig 3 ) . The Tibetan oxyHb SNP set also showed a similar pattern ( p = 0 . 102 and 0 . 046 for Tibetans and Sherpa , respectively; S4 Fig ) . However , when the EPAS1 SNP rs372272284 was excluded , no difference between the Hb- or oxyHb-associated SNPs and control SNPs was observed ( p ≥ 0 . 211; Fig 3 and S4 Fig ) . Thus , the overall frequency difference seemed entirely due to the large frequency differentiation of the EPAS1 SNP: the Hb-increasing allele frequency was 0 . 253 and 0 . 990 for Tibetans and CHB , respectively . GWAS SNP effect sizes have been shown to be correlated between European and East Asian populations [62] , implying that at least part of the SNPs identified in the large Hb GWAS in Europeans are also likely to be associated with Hb in our Tibetan sample . Therefore , we also tested for polygenic adaptation using SNPs identified in a large European GWAS [52]; of the 140 European GWAS SNPs , we used the 91 SNPs that were called in our data set . This set of SNPs did not include any EPAS1 SNPs , because the Tibetan EPAS1 haplotype is virtually absent outside Tibetan populations [63] . We found a trend towards lower frequencies of Hb-increasing alleles in both Tibetan and Sherpa , but this trend reached nominal levels of significance only in Sherpa ( p = 0 . 249 and 0 . 019 for Tibetans and Sherpa , respectively; Fig 3 ) . Consistent with the results of the pairwise test , neither the overdispersion test nor the outlier test for the high-altitude populations yielded results reaching nominal levels of significance ( poverdispersion = 0 . 695 and poutlier = 0 . 066 with no multiple test correction ) for oxyHb , or Hb ( poverdispersion = 0 . 201 and poutlier = 0 . 846 with no multiple test correction ) when the EPAS1 SNP was excluded ( Fig 3 , S4 Fig and S8 Table ) . Similar to the pairwise population test results , the outlier test was highly significant when the EPAS1 SNP was included ( p ≤ 0 . 0008; Fig 3 , S4 Fig and S8 Table ) . Using the Hb associated SNPs from the European GWAS , we again observed a trend toward lower genetic values in the high altitude populations , but it did not reach levels of statistical significance ( poverdispersion = 0 . 433 and poutlier = 0 . 110 with no multiple test correction ) . Therefore , these analyses do not provide evidence that alleles associated with lower Hb levels were selected for , except for the EPAS1 locus . Given that the EPAS1 SNPs explain a small fraction of the total variation in Hb levels ( 2 . 7% in our cohort ) , these results raise the question of whether unelevated Hb per se was the adaptive trait in Tibetans . In contrast to Hb and oxyHb , deoxyHb showed significant polygenic adaptation signals , with both the outlier and the pairwise difference tests ( poutlier = 0 . 023 and ppairwise = 0 . 002; Fig 4 , S4 Fig and S8 Table ) . This result is compatible with the lack of evidence for polygenic adaptation toward lower Hb in post-reproductive females because deoxyHb is not strongly correlated with deoxyHb ( r = 0 . 441 compared to that with oxyHb r = 0 . 874 ) . The alleles associated with higher deoxyHb in Tibetans were on average less common in Tibetans than in 1KGP CHB and the genetic values of deoxyHb in Tibetans and the Sherpa tended to be lower than those in 1KGP East Asians . Maximizing oxygen delivery while minimizing blood viscosity is likely to be beneficial in high-altitude environments; therefore , this advantage may underlie our signal of polygenic adaptation for lower deoxyHb . Based on 123 overlapping SNPs ascertained from a recently released large GWAS using the UK Biobank data [64] , we find a very strong signal for low pulse in the pairwise difference test ( ppairwise = 0 . 0001; S5 Fig and S8 Table ) . Results using 52 SNPs ascertained from our Tibetan GWAS also show a marginally significant deviation in the same direction ( ppairwise = 0 . 0620 ) . Interestingly , we do not see significant signals in either overdispersion or outlier test for Tibetans or Sherpa ( p > 0 . 05 ) ; instead , we see a strong signal in the outlier test for lowland East Asians toward high pulse ( poutlier = 0 . 001; S5 Fig ) . The most parsimonious interpretation of these findings is that selection favored higher pulse only in low altitude East Asians , although other explanations are also possible . We did not find a significant result for any polygenic adaptation test for SaO2 ( S8 Table ) . The GWAS of reproductive traits allowed us to identify candidate variants that are currently being selected for in the sampled Tibetan population . In our results , none of the most strongly associated variants with reproductive outcomes showed strong signals of selective sweeps . However , if these variants affected reproductive fitness in the past in addition to the current generation , we might expect signals of polygenic adaptation . Indeed , a number of reproductive traits showed strong signatures of polygenic adaptations based on the outlier test; the pairwise population difference test , which uses less information and hence is likely to be less powerful , gives broadly consistent results , although at lower levels of significance ( Fig 4 , Table 3 , S6 Fig and S8 Table ) . We see significant polygenic adaptation signals in several measures directly related to reproductive fitness . Interestingly , the significant signals are observed for both the viability ( e . g . the number of children born alive but died before 15 years; poutlier = 0 . 000 , ppairwise = 0 . 024 ) and the fertility fitness component ( e . g . the number of live births; poutlier = 0 . 002 , ppairwise = 0 . 002 ) . Furthermore , consistent with expectations , alleles increasing offspring mortality were selected against whereas those increasing offspring survival were positively selected for . A variable known to be directly linked to reproductive fitness , i . e . a woman’s age at her first childbirth , is also under selection , with earlier ages being advantageous , as expected . Twinning appears to have been selected against in Tibetan women . Although twinning may increase fitness , it is also associated with increased risks to mother and offspring due to limits on women’s ability to support adequate weight gain for two babies during the third trimester and to the lower birth weight of twins relative to singletons [65] , which in turn is associated with higher neonatal and infant mortality .
We identified several genome-wide significant associations with key physiological and fertility phenotypes in Tibetans ( Fig 2 and Table 2 ) , by analyzing new dense genome-wide variation data of over 1 , 000 indigenous inhabitants above 3 , 000 m in Nepal ( 2 , 982–4 , 052 m with mean of 3 , 630 m; S1 Table ) . Using genetic variants identified in our Tibetan GWAS , we found that several phenotypes showed signatures of polygenic adaptation towards better reproductive outcomes ( e . g . the number of livebirths ) ( Table 3 , S6 Fig and S8 Table ) . We also found evidence for polygenic adaptations for changes in pulse , possibly due to selection for higher pulse in low altitude East Asians that did not act on Tibetans and Sherpa . Surprisingly , we did not find clear evidence for polygenic adaptation towards low Hb in Tibetans beyond a link through the EPAS1 gene , even though we confirmed a correlation between low Hb and better reproductive outcomes . Because Hb concentration is a polygenic trait , these results raise the question of whether lower hemoglobin is causally related to higher reproductive fitness . The availability of reproductive history data in a population with little or no birth control offers unique opportunities for elucidating the adaptation process . Indeed , the ethnic Tibetan women sampled in this study have high birth rates ( the number of livebirths = 5 . 38 ± 2 . 79; mean ± 1 standard deviation ) and live in a mostly traditional society , where modern medical care , including in some regions contraception , has been introduced only very recently [48] . The reproductive data , collected in women who had largely completed their family size , allowed testing for a relationship between genetic or phenotypic variation and fitness differential . Genetic variation carrying well-established signals of selective sweeps , i . e . EGLN1 and EPAS1 SNPs , was not significantly associated with reproductive success probably due to low power: we estimated that the lowest selection coefficients that we had 80% power to detect were 6 . 6% and 7 . 4% at the EGLN1 and EPAS1 SNPs , respectively ( S7 Table ) . These selection coefficients are well above those proposed based on population genetics studies [37 , 61] . However , we did detect significant signals of polygenic adaptations using the SNPs identified in our GWAS of fertility variables . Importantly , alleles increasing survival variables or decreasing death variables were selected for ( Table 3 and S8 Table ) . Because the alleles influencing reproductive outcomes in Tibetans are common also at low altitude , we would not expect them to have changed systematically in frequency in one subset of populations , i . e . Tibetans . Therefore , an important implication of our findings is that the alleles increasing reproductive success in Tibetans interact with either high altitude environmental conditions or with other genetic variants that are common among Tibetans but not at low altitude . This scenario strongly supports the efforts to conduct studies of genetic and phenotypic diversity in diverse populations [66] living in their ancestral environment , despite enormous logistical challenges . An attenuated erythropoietin and Hb concentration response to hypobaric hypoxia is a hallmark phenotype of the “Tibetan pattern” of high-altitude adaptations , which is markedly different from that of Andean highlanders [32 , 67 , 68] . The low prevalence among Tibetans of diseases associated with elevated Hb concentration , such as chronic mountain sickness [69] , and a signal of selective sweep in the EPAS1 gene [33 , 34] have led to the hypothesis that unelevated Hb is adaptive in Tibetan highlanders [70]; this hypothesis was also substantiated by the correlation between low Hb and better reproductive outcomes in our Tibetan sample [48] . Our GWAS provides the first genome-level support for the association between the Tibetan EPAS1 haplotype and low oxyHb , which correlates highly with total Hb . Interestingly , the association was stronger for oxyHb than for total Hb ( Table 2 and S4 Table ) , while it was not significant for deoxyHb ( p = 0 . 883 for rs372272284 ) . This observation raises the possibility that it is the oxygen-carrying portion of total Hb that drives the well-replicated association between EPAS1 SNPs and Hb . We also found that SNPs associated with Hb did not show polygenic adaptation signals in our Tibetan sample , if the EPAS1 SNP was excluded from the analysis ( Fig 3 ) . Intriguingly , the Sherpa , who are closely related to other Tibetan populations and also have unelevated Hb levels [41 , 67 , 71] , show a nominally significant trend towards lower frequencies of the Hb-increasing alleles in one of the two polygenic adaptation tests ( p = 0 . 019 without multiple test correction ) . Based on our estimate of 0 . 386 g/dL per allele , and a mean allele frequency difference of 0 . 743 between high and lowlanders , we calculated that the EPAS1 SNPs can explain 52% of the 1 . 1 g/dL difference reported in [72] between Tibetan and Han Chinese women in the same age range . In our sample , the EPAS1 SNP explains only 2 . 7% of inter-individual variation in Hb: therefore , almost all within-population ( 97 . 3% ) as well as a substantial portion of between-population ( 48% ) variation remains unexplained . Several scenarios could account for these results . Incomplete power in the Tibetan GWAS and/or in the polygenic adaptation tests could underlie the lack of clear evidence for polygenic adaptation for lower Hb levels , although we had sufficient power to detect polygenic adaptation signals for several other traits in the same samples . One possibility is that post-reproductive Hb levels are a poor proxy for the levels while women are reproductively active . Some , but not all , studies of Tibetans find an increase in Hb concentration with age [73–75] , but this does not imply that the relationship between genotype and phenotype also changes with age ( especially if age is used as a covariate in mapping , as done here ) . The lack of evidence supporting low Hb as the selected trait in Tibetans stands in stark contrast with the strong selective sweep signal at EPAS1 and with the significant evidence for polygenic adaptations toward lower deoxyHb . This finding raises the question of whether unelevated Hb was the true target of selection in Tibetans rather than a mere correlate of the true adaptive trait . This scenario would be consistent with the observed correlation between low Hb and better reproductive outcomes because pleiotropy can induce a non-causal association between phenotypes . A recent study showed that the same EPAS1 SNP that is associated with Hb and other hematological traits is also associated with uric acid levels [38] , suggesting that indeed SNPs in EPAS1 , a transcription factor with dozens of target genes , may affect multiple , seemingly unrelated phenotypes . Interestingly , the peak of our association signal for oxyHb at EPAS1 spans active enhancer ( H3K27Ac ) marks in human umbilical endothelial cells , as detected by the ENCODE project [76] , pointing to gene regulatory role in the endothelium . Therefore , it could be speculated that the SNPs that influence variation in oxyHb/Hb levels also affect EPAS1 expression in the endothelium with effects on vascularization , vasoconstriction , vasodilation and possibly beneficial effects in oxygen delivery at high altitudes . These findings suggest that the WHO altitude-adjusted elevated hemoglobin cut-off for detecting iron-deficiency anemia [77] may be inappropriate for use among Tibetan women , a result of this work that has public health implications and that warrants further research . The lower genetic values for pulse in Tibetans compared to low altitude East Asians , coupled with the correlation between lower pulse and better reproductive outcomes in Tibetans , suggest an important role for cardiac function in pregnancy at high altitude . Intriguingly , tests of polygenic adaptation that use data from worldwide populations are consistent with selection favoring higher pulse in low altitude East Asians , but not in the closely related populations at high altitude . There is some prior evidence for selective events that took place in low altitude , but not high altitude East Asians . For example , the well-known selective sweep signal at the ADH locus [78] in low altitude East Asians is not shared with Tibetans: the derived allele at the nonsynonymous SNP rs3811801 is very common in Han Chinese and Japanese ( CHB = 0 . 59 and JPT = 0 . 70 ) but relatively rare at high altitude ( Tibetan = 0 . 08 and Sherpa = 0 . 07 ) . A similar pattern is seen for the rs1800414 derived allele at the OCA2 gene ( CHB = 0 . 59 and JPT = 0 . 57 versus Tibetan = 0 . 06 and Sherpa = 0 . 16 ) . Therefore , the observed shift towards higher genetic values for pulse at low altitude could be the result of a selective event that similarly affected only low altitude populations . However , the correlation between low pulse and better reproductive outcomes in Tibetans suggests that low pulse is adaptive ( rather than neutral ) at high altitude and raises the possibility that lower pulse was selected for when ancestral low altitude populations moved to high altitude . Our GWAS of fertility phenotypes discovered three genome-wide significant associations ( Table 2 and S1 Fig ) . Those signals lie in or near genes of potential biological relevance . First , the association peak for the number of pregnancies and of livebirths is located within an intron of the CCDC141 gene ( Fig 2 ) , which is expressed in the heart and had been linked to a rare form of hypogonadotropic hypogonadism [79] . This gene is an immediate neighbor of the TTN ( titin ) gene , which codes for a major component of cardiac muscle and has been linked to idiopathic dilated and peripartum cardiomyopathy and cardiac remodeling [80 , 81] . Genetic variants within 6 kb from our association peak were reported to be associated with cardiac phenotypes , such as heart rate [53 , 82] . Although our GWAS signals were not associated with pulse , we hypothesize that they influence heart function , which in turn may affect pregnancy outcomes in the extreme high-altitude environments . The observed negative correlation between pulse and the number of livebirths is consistent with this idea . Second , the top SNP in chromosome 14 associated with the number of stillbirths is 99 kb away from the PAPOLA gene encoding a poly-A tail polymerase that affects mRNA stability and nuclear export . Intriguingly , the product of this gene is inhibited by cordycepin , an adenosine analog ( 3’ deoxyadenosine ) , found in a fungus , “Yartsa gunbu” or Cordyceps sinensis , which is native to the highlands of Nepal and Tibet . Cordycepin is known to interfere with important biochemical and molecular processes , such as purine biosynthesis , DNA/RNA synthesis and mTOR ( mammalian target of rapamycin ) signaling transduction ( reviewed in [83] ) . Therefore , cordycepin exposure during pregnancy could have negative effects on reproductive outcomes . Harvest of this fungus for sale primarily in China is a major source of household revenue in the Gorkha district , from where about one third of our participants were recruited . Although it is not a species consumed by ethnic Tibetan women in this region , our results raise the possibility that the PAPOLA SNPs may affect the stillbirth phenotype by interacting with an exposure to C . sinensis during pregnancy . An alternative and equally likely explanation is that these SNPs influence reproductive outcomes through mechanisms not involving cordycepin exposure , for example by affecting mRNA levels of key genes involved in inflammatory processes , as suggested in knockdown experiments of the PAPOLA gene [84] , or through mechanisms involving other nearby genes . This study was designed to extend the genetic study of human local adaptation beyond selective sweeps and candidate gene associations , by collecting genotype and physiological phenotype and reproductive history data for a large group of indigenous high-altitude Tibetan women in Nepal . Using this data set , we successfully identified several new genome-wide associations and signatures of polygenic adaptations . Our sample size of 1 , 000 participants is remarkably large for the genetic study of populations living in remote locations in a traditional society , but we acknowledge that is rather small for a modern-day GWAS . The census population size of ethnic Tibetans of villages in this region set the ultimate constraint on our sample size , which was obtained by recruiting virtually all inhabitants fitting our inclusion criteria . Despite this constraint , this study shows the necessity to study phenotypes of locally adapted populations in their native environments to correctly identify the adaptive phenotypes . With ever increasing throughput to generate genetic and phenotypic variation data , in-depth phenotyping of potentially adaptive features will help better understand how Tibetans and other populations living in extreme environments have adapted to their habitats .
The study protocol was approved by: the University Hospitals Institutional Review Board , University Hospitals of Cleveland ( protocol no . 12-15-27 ) , the Nepal Health Research Council , Kathmandu , Nepal ( protocol no . 38/2011 ) , the Oxford Tropical Research Ethics Committee , Oxford , UK ( protocol no . 23–11 ) , the Dartmouth College Committee for the Protection of Human Subjects ( protocol no . 23374 ) and the Human Research Protection Office , Washington University in St . Louis ( protocol no . 201202114 ) . A written informed consent was signed by each participant . A total of 1 , 008 ethnic Tibetan participants were recruited from high-altitude villages in Mustang and Ghorka districts in Nepal in the spring and summer of 2012 . All participants were women of age 39 or older and lifelong residents above 3000 m of altitude . The study communities in Nepal lie on the southern aspects of the Tibetan Plateau . Although they are citizens of Nepal , local people speak Tibetan dialects , practice forms of religion and social organization akin to those across the Tibetan Plateau , and retain the characteristic agro-pastoral and trading mode of subsistence common among highland Tibetans [48] . An additional 103 Sherpa participants were recruited from high-altitude villages in the Khumbu district in Nepal in the summer of 2014 . Most of the Sherpa participants were women of age 39 or older . We collected saliva samples of husbands and children for 12 of them . Saliva samples were collected in the field using OG-500 Oragene DNA collection kits ( DNA Genotek Inc . , Otawa , ON , Canada ) and genomic DNA ( gDNA ) were extracted using the prepIT-L2P reagents ( DNA Genotek Inc ) following the manufacturer’s protocol . Blood hemoglobin concentration ( Hb ) , percent arterial blood oxygen saturation ( SaO2 ) , and pulse rate ( pulse/minute ) were measured altogether using a non-invasive device Masimo Pronto-7 © ( Masimo Corporation , Irvine , CA ) as described in Cho et al . [48] . Two additional phenotypes , oxygenated and deoxygenated hemoglobin concentrations ( oxyHb and deoxyHb , respectively ) , were calculated from Hb and SaO2 as follows: oxyHb = Hb × SaO2 / 100 and deoxyHb = Hb–oxyHb . For each participant , an interview session was held to retrieve detailed reproductive history as well as to collect other potential covariates . A summary of the Tibetan samples and their phenotype measurements are presented in Table 1 . Detailed description of the Tibetan samples , the phenotype and covariate data collection was published in Cho et al . [48] . We generated new genome-wide genotype data for a total of 1 , 104 individuals indigenous to the high-altitude regions in the Himalayas in Nepal , including 1 , 001 ethnic Tibetans from the present study and 103 Sherpa ( S1 Table ) . Array genotyping was performed in two phases . First , all Tibetan individuals were genotyped on 301 , 299 biallelic markers using the customized Illumina HumanCore-12 v1 . 0A array , which includes probes for additional 2 , 553 markers from 19 genomic loci presumed adaptive in Tibetans including the EPAS1 , EGLN1 , HIF1A and NOS2 genes . Then , a subset of 344 unrelated Tibetans from the present study and all 103 Sherpa individuals were genotyped on 716 , 503 markers using the Illumina OmniExpress-24 v1 . 0 array to obtain denser genome-wide variation data . For each array platform , genotypes were called using the genotyping module in the Illumina Genome Studio with default parameters ( GenCall score threshold 0 . 15 ) . Previously defined clusters , downloadable from the Illumina website , were applied for genotype calling . For the 2 , 553 custom markers we added to the HumanCore array , we retrieved intensity data from the Illumina Genome Studio and performed genotype calling using the OptiCall v0 . 6 . 4 [85] . For 344 Tibetans genotyped on both Illumina platforms , we used genotype calls from the HumanCore array for the overlapping 253K markers . Genotype calls from the two platforms were highly concordant , with the average 99 . 98% concordance . We separately genotyped two non-synonymous SNPs in the EGLN1 gene , rs12097901 and rs188966510 , in the set of 344 unrelated Tibetans . We used Epicenter FailSafe PCR system with the manufacturer’s recommended condition in buffers G and H , instead of using standard TAQ polymerases . We generated a 1 , 025 bp PCR fragment in an 11 ul reaction volume using a previously published primer pair PHD2-X1F ( CCCCTATCTCTCTCCCCG ) and PHD2-X1R ( CCTGTCCAGCACAAACCC ) [86] . These PCR products were sequenced using BigDye Terminator v3 . 1 cycle sequencing kit and the PHD2-X1F primer in an Applied Biosystems 3730XL DNA Analyzer . In a few cases where initial amplification failed , samples were diluted 4x in water , which in most cases allowed successful subsequent amplification . Genotypes were scored manually from chromatograms . We generated novel whole genome sequence data for 18 Sherpa and 35 Tibetans from the present study , all from Nepal . Seventeen individuals were sampled with known familial relationships ( four Tibetans mother-daughter duos and three Sherpa parents-offspring trios ) , and sequenced to high-coverage ( around 20x autosomal coverages ) to generate high quality phased genome sequences . The remaining 36 individuals were chosen to be unrelated and sequenced to low-coverage targeting 5x autosomal coverage . For Sherpa , we began with 172 individuals , including 103 newly genotyped in this study and 69 previously published [41] , and chose a subset of 101 unrelated individuals allowing first cousins . Coefficients of relatedness were calculated using PLINK v1 . 07 [87] . Then , we estimated population structure in these unrelated Sherpa , together with 30 Tibetans from near Lhasa [42] and 103 1KGP CHB , using an unsupervised genetic clustering algorithm in ADMIXTURE v1 . 22 [88] . Using estimates from K = 2 , we chose 51 Sherpa with > 95% of their ancestry from a component enriched in Sherpa and Tibetans ( the remaining portion come from an ancestry representing CHB-related low altitude East Asians ) . Among them , we chose three pairs of couples with their offspring and 9 additional unrelated individuals for high- and low-coverage sequencing , respectively . For Tibetans , we ran ADMIXTURE with K = 3 in a supervised mode , with 103 1KGP CHB , 103 1KGP GIH ( Gujarati Indians in Houston , Texas ) and the 51 unrelated Sherpa as three reference groups . Pairwise relatedness was then calculated with the ADMIXTURE output using the RelateAdmix v0 . 08 , controlling for population structure due to admixture [89] . Among individuals with minimum South Asian ancestry ( < 1% ) , represented by GIH , we chose four pairs of mother-daughter duos of Tibetans from the present study and 27 unrelated individuals for high- and low-coverage sequencing , respectively . Single-barcoded libraries for Illumina sequencing were constructed using the TruSeq library preparation kit . Libraries were pooled into multiple batches and sequenced in the Illumina HiSeq 2500 and 4000 machines for paired-end ( PE ) 100 and 125 bp designs ( S1 Table ) . Sequence reads were demultiplexed with no mismatch in 6-bp barcode sequence allowed . Reads were mapped to the human reference genome sequences ( hg19 ) downloaded from http://hgdownload . soe . ucsc . edu/goldenPath/hg19/chromosomes/ , using BWA backtrack v0 . 7 . 4 with -q15 option [90] . PCR and optical duplicate reads were marked using Picard tools v1 . 98 ( http://broadinstitute . github . io/picard/ ) and were excluded from further analysis . Local realignment around indels and base quality score recalibration were performed using the GenomeAnalysis ToolKit ( GATK ) v2 . 8–1 , following the best practice pipeline [91–93] . Finally , analysis-ready BAM files for variant discovery and genotype calling were produced using Samtools v1 . 2 [94] by filtering out reads with Phred-scaled mapping quality lower than 30 . LD-aware variant and genotype calling was performed using the GotCloud pipeline [95] with default parameters . The analysis-ready BAM files of 53 newly sequenced individuals and 6 previously reported ones , four Sherpa and two Nepali Tibetans [41 , 43] , were provided to the pipeline together . We performed genotype imputation of Tibetan and Sherpa samples , which were array-genotyped either in the present or in our previous study [41] ( S1 Table ) . For each array genotyping platform , low quality markers and samples were filtered out by applying the following filters: per-marker missing rate ≤ 0 . 05 , Hardy-Weinberg equilibrium ( HWE ) p-value ≥ 0 . 00001 and per-individual missing rate ≤ 0 . 03 . Strand-ambiguous ( A/T and G/C ) SNPs were removed and only SNPs in autosomes or X chromosome were retained for imputation . The filtering process was performed using PLINK v1 . 90 [96] . Genotype imputation was performed for each set of samples separately using IMPUTE2 v2 . 3 . 2 [46] . We used both our phased genotype calls of 59 high-altitude samples and the 1KGP phase 3 reference data set , downloadable from https://mathgen . stats . ox . ac . uk/impute/1000GP_Phase3 . html , as imputation references by merging them with “-merge_ref_panels” flag in IMPUTE2 . For other parameters , we used default values set by the program . Following imputation , genotypes with posterior probability ≥ 0 . 9 were accepted . Genotypes were assumed to be missing if none of three possible genotypes reached posterior probability threshold of 0 . 9 . Then , we conducted an additional round of quality control by removing SNPs with missing rate higher than 0 . 05 or HWE p-value smaller than 10−6 . Among 1 , 001 successfully genotyped and imputed Tibetan women , 991 individuals were included in our genome-wide association analysis ( GWAS ) . Four individuals were excluded from the analysis because they were born below 3 , 000 m . Another individual was excluded from the analysis was a genetic outlier who clustered with individuals from the Indian subcontinent . The other five were excluded either because they had inconsistent reproductive record or because they were nuns who became celibate during their reproductive years . For physiological phenotypes , we chose relevant covariates by performing a stepwise model selection , allowing removal of a single covariate each step if likelihood ratio test ( LRT ) p-value obtained from the “lrtest” function in the R “lmtest” library was bigger than 0 . 05 . The final sets of chosen covariates for physiological covariates are listed in S3 Table . For fertility phenotypes , we used an a priori chosen set of four covariates: age , subdistrict , use of contraception and “continuously married ( CM ) ” status . Use of contraception was categorized into three classes: never used , previously used , and currently in use . “Continuously married” status is a binary variable defined as being in a marital relationship throughout the ages of 25 and 40 . It includes two who had experienced less than two years of gap before re-marriage following divorce or death of the husband . Table 1 presents a summary of these covariates . A full list of GWAS phenotypes and their description are provided in S3 Table . GWAS was performed using GEMMA v0 . 94 . 1 [47] . Univariate linear mixed model ( LMM ) as implemented in GEMMA was used to control for both population structure and genetic relatedness [47] . For each phenotype , we first removed individuals with no information on either the focal phenotype or its associated covariates . Second , we kept SNPs with maf ≥ 5% for the chosen subset of individuals . Third , the standardized genetic covariance matrix was calculated from this data set and was used for LMM . Last , GWAS was run controlling for the above covariates . For continuous and count data , we provided raw phenotype data together with covariates to the program . For the binomial data , we fitted a binomial regression model using the “glm” function in R , calculated the difference between the observed odds and the odds of the fitted value , and used this residual as a GWAS phenotype . LRT p-values from GEMMA were used to assess significance of genetic association . P-values of the full and subsample sets were highly correlated for each fertility phenotype ( Pearson r = 0 . 36–0 . 74 with p < 10−15 for–log10 transformed p-values ) . To detect signatures of polygenic adaptation , we investigated systematic changes in allele frequencies of SNPs associated with each phenotype . For all of Tibetan GWAS phenotypes , we first took all SNPs with p ≤ 10−4 and lumped them into peaks by allowing maximum inter-SNP distance of 200 kb . Finally , we chose one SNP with the smallest association p-value for each peak to retrieve a set of independently associated SNPs . We also retrieved a set of SNPs associated with blood hemoglobin level ( Hb ) using summary statistics from a published large-scale GWAS meta-analysis [58] . To obtain a list of independent markers , we first confined markers to those overlapping with our Tibetan data and applied a more stringent cutoff of p ≤ 10−5 . Then , we removed SNPs in LD: for each pair of SNPs with r2 > 0 . 2 in 1KGP CEU , we removed one with larger association p-value . After retrieving phenotype-associated SNPs with their effect size , we first compared mean frequency difference of trait-increasing alleles between Tibetans and 1KGP CHB . Following [13] , we sampled 10 , 000 sets of random SNPs , where each set contained an equal number of SNPs as the GWAS SNPs matched one-to-one by mean minor allele frequency in bins of size 0 . 02 . The empirical distribution of mean frequency difference of trait-increasing alleles was compared to the observed value from the GWAS SNPs and the empirical one-sided p-value was calculated as the proportion of random SNP sets with their mean allele frequency difference equal to or more extreme than the observed one . We also looked into comprehensive signatures of polygenic adaptation using a machinery introduced by [14] . It requires as an input a set of SNPs associated with the target phenotype together with their allele frequency and effect size estimate . For each population , a “genetic value” of the target phenotype is calculated as a weighted sum of population allele frequency over the GWAS SNPs with the effect size as a weight . Then , the calculated genetic value is used for a set of tests asking i ) if the GWAS SNPs are collectively more differentiated between populations than the matched random SNPs are , ii ) if the direction of allele frequency differentiation is more consistent in GWAS SNPs than in matched random SNPs , iii ) if the genetic value is correlated with an environmental variable over populations , or iv ) if a regional group or a population’s genetic value is away from the expected value by the genetic value of the other populations . For this , we used allele frequency of 26 populations in the 1KGP phase 3 data set overlapping with the Tibetan data . We first sampled random SNPs matching each of the GWAS SNPs by minor allele frequency bin of size 0 . 02 in the GWAS population and by the B-value bin of size 100 ( values ranging from 0 to 1 , 000 ) [97] . We sampled up to several thousands of random SNPs per GWAS SNP to obtain around 100 , 000 random SNPs in total . These random SNPs were used for calculating the genetic covariance matrix of populations and for generating 5 , 000 sets of matched random SNPs . To estimate the strength of positive selection necessary to generate a significant association between the fertility count phenotypes and genotype in our sample of the unrelateds , we assumed a simple additive model . That is , genotypes with 0 , 1 and 2 adaptive alleles , with population frequency p , have the mean absolute fitness W0 , W0 ( 1+s ) and W0 ( 1+2s ) . Using the observed mean phenotype value , Wm , we can get the per-allele effect size sW0 as a function of s , Wm and p: sW0=sWm1+2sp Then , the effect size was standardized to the unit of standard deviation , using the observed standard deviation of the phenotype . For the standardized effect size , which is a function of selection coefficient s , the statistical power to detect association was calculated using the “pwr . r . test” function in the R package “pwr” . | The adaptations to high altitude environments in Tibetan populations have long been highlighted as an important case study of adaptive evolution in our species . Recent genetic studies found two genes , EGLN1 and EPAS1 , the genetic variants in which were swept to high frequency in Tibetans due to strong positive natural selection . However , it still remains unclear if and how these and other genetic variants are connected to adaptive phenotypes and ultimately to fitness advantage . In this study , we collected genotype and phenotype information of 1 , 000 ethnically Tibetan women from the high Himalayan valleys in Nepal . Using both genome-wide association analysis and test for polygenic adaptations , we show that natural selection systematically altered frequency of alleles associated with reproductive outcomes to the direction of increasing fitness . That is , alleles associated with more pregnancies and live births , as well as those associated with lower offspring mortality , were under positive selection . Omitting the EPAS1 haplotype under selective sweep , the other variants associated with lower hemoglobin did not collectively show a clear signal for polygenic adaptation . Our study highlights the polygenic nature of human adaptive traits . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"genome-wide",
"association",
"studies",
"medicine",
"and",
"health",
"sciences",
"maternal",
"health",
"obstetrics",
"and",
"gynecology",
"population",
"genetics",
"alleles",
"ethnicities",
"women's",
"health",
"pregnancy",
"genome",
"analysis",
"population",
"biology",
"tibetan",
"people",
"genetic",
"loci",
"people",
"and",
"places",
"phenotypes",
"natural",
"selection",
"evolutionary",
"processes",
"genetics",
"population",
"groupings",
"biology",
"and",
"life",
"sciences",
"genomics",
"evolutionary",
"biology",
"computational",
"biology",
"human",
"genetics"
] | 2018 | Detecting past and ongoing natural selection among ethnically Tibetan women at high altitude in Nepal |
MicroRNAs ( miRNAs ) are post-transcriptional regulatory RNAs that can modulate cell signaling and play key roles in cell state transitions . Epstein-Barr virus ( EBV ) expresses >40 viral miRNAs that manipulate both viral and cellular gene expression patterns and contribute to reprogramming of the host environment during infection . Here , we identified a subset of EBV miRNAs that desensitize cells to B cell receptor ( BCR ) stimuli , and attenuate the downstream activation of NF-kappaB or AP1-dependent transcription . Bioinformatics and pathway analysis of Ago PAR-CLIP datasets identified multiple EBV miRNA targets related to BCR signal transduction , including GRB2 , SOS1 , MALT1 , RAC1 , and INPP5D , which we validated in reporter assays . BCR signaling is critical for B cell activation , proliferation , and differentiation , and for EBV , is linked to reactivation . In functional assays , we demonstrate that EBV miR-BHRF1-2-5p contributes to the growth of latently infected B cells through GRB2 regulation . We further determined that activities of EBV miR-BHRF1-2-5p , EBV miR-BART2-5p , and a cellular miRNA , miR-17-5p , directly regulate virus reactivation triggered by BCR engagement . Our findings provide mechanistic insight into some of the key miRNA interactions impacting the proliferation of latently infected B cells and importantly , governing the latent to lytic switch .
Epstein Barr virus ( EBV ) is a human gamma-herpesvirus that infects >90% of adults worldwide . Primary infection is often asymptomatic or presents as infectious mononucleosis . In immunocompromised individuals , the virus is linked to post-transplant lymphoproliferative disease ( PTLD ) and hematological malignancies including Burkitt’s ( BL ) lymphoma , Hodgkin lymphoma , and AIDS-related lymphoma; cancers of epithelial origin such as gastric and nasopharyngeal carcinomas are also associated with EBV infection [1 , 2] . Following EBV transmission through the oral cavity and subsequent infection of naïve B cells , EBV co-opts multiple aspects of normal B cell activation that induces cell proliferation , initiates differentiation programs , and can drive infected B cells through germinal center ( GC ) reactions to establish latency in the memory B cell compartment [3–5] . Periodic virus reactivation can occur throughout life-long infection of the host and is thought to help maintain the pool of latently infected cells [5 , 6] . Like all herpesviruses , EBV has both latent and lytic replication phases , and key to the success of long-term persistence is the ability to navigate between these phases . EBV encodes over 85 open reading frames ( ORFs ) and several non-coding RNAs ( ncRNAs ) which are temporally and differentially expressed throughout the viral life cycle . At least three distinct latency programs ( I , II , III ) have been described for EBV , distinguished by viral gene expression patterns [7] . The latency III profile consists of nine latent genes ( EBNA1 , EBNA2 , EBNA3A-C , EBNA-LP , LMP2A/B , LMP1 ) as well as ncRNAs such as the EBERs and 25 precursor microRNAs ( pre-miRNAs ) which are processed into >40 mature viral miRNAs . Three EBV BHRF1 pre-miRNAs ( miR-BHRF1-1 , -2 , -3 ) , encoded in the BHRF1 locus , are highly expressed following de novo infection , lytic replication , and during latency III when the Cp or Wp promoters are active , such as in lymphoblastoid cell lines ( LCLs ) derived in vitro [8] . BHRF1 miRNAs can also be detected in some EBV+ B cell tumors [9] . With the exception of miR-BART2 , which is encoded anti-sense to EBV BALF5 , the remaining EBV miRNAs are clustered within the BART region and are detectable at varying levels in all EBV infection stages ( reviewed in [10] ) . While exact functions for EBV miRNAs continue to emerge , previous studies demonstrate that these small , viral ncRNAs act akin to their cellular counterparts to post-transcriptionally regulate gene expression via interactions with both viral and cellular RNA targets . EBV and other g-herpesvirus miRNAs are known to regulate multiple cell signaling pathways , including those associated with innate and cell-mediated anti-viral immune responses such cytokine and interferon signaling [11–14] . Specific examples include EBV miR-BHRF1-3 , which downregulates the chemokine CXCL11 in primary lymphomas [9]; miR-BART15 can target NLRP3 to alter inflammasome activation [15]; EBV miR-BART16-5p targets CREB-binding protein [16] and can disrupt type I IFN signaling [14]; EBV miR-BHRF1-2-5p targets IL1R1 , encoding the major IL-1 receptor , thereby dampening IL-1 signaling [11] . In addition to IL1R1 , we recently reported that the EBV BHRF1-2 miRNAs target other IL-1 signaling components such as SOS1 , a Ras GDP/GTP exchange factor , and PLCG1 , encoding phospholipase C gamma 1 that contributes to receptor-mediated tyrosine kinase signal transduction [11] . Both of these components are integral to many other signaling pathways , such as B cell receptor ( BCR ) signaling . Furthermore , published miRNA targetome studies suggest that EBV miRNAs target several cellular components related to B cell activation [17–19] , leading us to hypothesize that the EBV miRNAs might functionally regulate signal transduction through the BCR . Aberrant BCR signaling is a hallmark of cancer progression; somatic mutations in cellular components that result in constitutively active or tonic BCR signaling contribute to enhanced proliferation of malignant B cells in cancers such as lymphoma [20 , 21] . BCR signaling is indispensable for normal B cell activation and differentiation . Engagement of the BCR occurs predominantly through antigen triggers that induce phosphorylation of cytoplasmic CD79a and CD79b immune receptor tyrosine-based activation motifs ( ITAMs ) by Src family of tyrosine kinases , thereby recruiting signalsome components such as Syk , Btk , Vav guanine exchange factors , and Grb2 and BLNK adaptor proteins to relay signals downstream [22] . Induction of BCR signaling subsequently results in activation of multiple transcription factors , such as NF-kappaB and Jun , which in turn induce genes that participate in B cell proliferation and survival . In the context of EBV infection , antigenic stimulation of the BCR can initiate plasma cell differentiation , which triggers the switch from latency to lytic replication [5 , 23 , 24] . Notably , EBV encodes viral proteins that can provide surrogate BCR survival signals thought to be necessary for establishing and maintaining persistent infection . In latency III cells , EBV-encoded latent membrane protein 2A ( LMP2A ) recruits Src and Syk tyrosine kinases via an ITAM within its cytoplasmic N-terminal domain , thereby mimicking the BCR [25] . LMP2A has growth transforming properties and may functionally replace BCR signals in B cells lacking an intact BCR [25–27] . Recent proteomic studies provide evidence that additional EBV-encoded factor ( s ) target the BCR complex for proteasomal degradation during entry into the lytic cycle [28] . While roles for EBV proteins in modulating BCR signal transduction have been described , roles for EBV miRNAs are not yet known . In this study , we sought to determine whether EBV miRNAs could influence BCR signaling and to subsequently elucidate the molecular mechanism ( s ) by which this may occur by determining cellular targets involved . To understand how the identified EBV miRNA target interactions might be relevant to the EBV life cycle , we examined gain and loss of function outcomes for specific EBV miRNAs in the growth of latently infected B cells as well as consequences for lytic reactivation triggered through surface Ig cross-linking .
To determine if EBV miRNAs could functionally regulate signaling through the BCR , we screened individual viral miRNAs in EBV-negative BJAB BL cells which express surface IgM and are responsive to antigenic triggering of the BCR using antibodies against IgM ( Fig 1 ) . BJAB cells stably expressing a NF-kappaB luciferase reporter ( BJAB-NFkB-GL4 . 32 ) were transduced with EBV miRNA expression vectors and then , cells were treated with anti-IgM to ligate the BCR and activate downstream signaling . We initially tested all three BHRF1 miRNAs , miR-BART2 , four BART miRNAs from BART Cluster 1 ( miR-BART1 , 3 , 5 , and 6 ) and four BART miRNAs from Cluster 2 ( miR-BART8 , 11 , 14 , and 18 ) ( Fig 1A ) . Notably , expression of six miRNAs ( BHRF1-2 , BART1 , BART2 , BART8 , BART11 , and BART18 ) significantly impacted the amplitude of the response to BCR stimulation as measured by reduced NF-kappaB activity . To examine whether basal NF-kappaB activity might be affected by EBV miRNA expression , we also tested several miRNAs in BJAB-NFkB-Luc cells [11] that constitutively express an internal control renilla luciferase in addition to a firefly luciferase NFkB reporter ( Fig 1B ) . We included BART9 and BART17 in these subsequent experiments based on published miRNA targetome studies indicating these two miRNAs might target the BCR signaling pathways [17 , 18] . In general , basal NFkB activity was not affected , while addition of anti-IgM led to a ~5-fold increase in NFkB reporter activity in pLCE control cells ( Fig 1B ) . Consistent with our initial experiments , BHRF1-2 and BART2 miRNAs reproducibly attenuated BCR signaling responses; we additionally observed decreased NFkB activity in the presence of BART9 and BART17 miRNAs ( Fig 1B ) . Together , these experiments identified eight EBV miRNAs that significantly reduced NFkB responses initiated through the BCR . We have recently demonstrated that the EBV BHRF1-2 miRNAs can reduce NF-kappaB activation in response to IL-1 cytokines [11] . There is extensive cross-talk between cytokine , BCR , and toll-like receptor ( TLR ) signaling pathways , leading us to consider the possibility that the BHRF1-2 miRNAs may be inhibiting core NF-kappaB components or other factors common amongst these pathways . We therefore examined the TLR4 signaling response by treating cells with lipopolysaccharide ( LPS ) . Compared to control cells , the presence of BHRF1-2 miRNAs had no impact on LPS-mediated NF-kappaB activation ( Fig 1C ) . These results indicate that activity of the BHRF1-2 miRNAs is directed more towards IL-1 and BCR pathways . Engagement of the BCR also activates JNK ( c-Jun N-terminal kinases ) and p38 , triggering MAPK signaling , and regulating activity of AP1 ( Activator protein 1 ) family transcription factors such as JUN and ATF2 [29 , 30] . Previously , miR-BART18-5p was shown to target the MAP3K2 ( MAP kinase kinase kinase 2 ) 3’UTR , suggesting that downstream transcription mediated through ATF2 and c-Jun may be affected [31] . Thus , to determine if AP1-dependent transcription is indeed influenced by EBV miRNAs , we generated BJAB cells stably expressing an AP1 luciferase reporter ( BJAB-AP1-Luc ) and measured responses to BCR cross-linking following 6 hr anti-IgM stimulation . Expression of the BHRF1-1 , BHRF1-2 , BART14 , or BART18 miRNAs significantly attenuated AP1 activation ( Fig 1D ) . We also tested 18 hr anti-IgM stimulation and observed significantly reduced AP1 activity in the presence of the BHRF1-2 miRNAs ( Fig 1E ) , indicating these miRNAs are highly involved in controlling BCR responses . Expression levels for individual EBV miRNAs in transduced BJAB cells are shown in Fig 1F . Together with the NF-kappaB reporter assays , these experiments demonstrate that multiple EBV miRNAs ( including BHRF1-2 , BART2 , BART9 , BART17 , and BART18 ) can functionally attenuate signaling events initiated through BCR engagement . As BJAB cells are not infected with EBV , these results further show that desensitization to BCR stimuli occurs through viral miRNA actions on the host cell , presumably through the partial inhibition or complete silencing of cellular targets . In order to identify the molecular mechanisms by which EBV miRNAs attenuate BCR signaling , we queried published B cell Ago PAR-CLIP datasets for experimentally derived , protein-coding mRNA targets of EBV miRNAs , focusing specifically on cellular targets for EBV BART2 , BART9 , BART17 , BART18 , and BHRF1-2 miRNAs . Combined datasets included EBV B95-8 LCLs ( expressing BHRF1-2 and BART2 miRNAs , among others ) , two EBV wild-type LCLs ( expressing all EBV miRNAs ) , and one EBV+/KSHV+ PEL ( expressing all 22 BART miRNAs ) [17 , 18 , 32] . 3 , 501 target 3’UTR interactions ( representing 1 , 891 unique human genes ) with 6mer ( nt 2–7 ) or greater seed match sites were extracted for the five EBV miRNAs and subsequently , analyzed for genes with established roles in BCR signaling using a manually curated list from six combined public resource pathway collection databases ( Reactome , Panther Pathways , NDeX , DAVID 6 . 8 ( KEGG and Biocarta ) , PathCards , and NetPath ) [33–38] . Based upon the literature , we also included INPP5D ( encodes the co-stimulatory phosphatase SHIP1 ) , PRDM1 ( encodes BLIMP1 , a master regulator of B cell differentiation ) , and PAG1 ( encodes a phosphoprotein that associates with Lyn and/or Fyn [39] ) . A total of 54 EBV miRNA targets were associated with BCR signaling including genes directly involved in Signaling by the B cell Receptor ( Reactome pathway R-HSA-983705 ) and involved in B cell activation ( Panther pathway P00010 ) ( Fig 2A ) . Targets include the SH2-domain containing adaptor Growth factor receptor-bound protein 2 ( GRB2 ) and Grb2 binding partners ( SOS1 ) that coordinate signaling downstream of the BCR to facilitate activation of Ras , MAPK , PI3K , and indirectly , NF-kappaB [40 , 41] as well as the RAC1 ( Ras-related C3 botulinum toxin substrate 1 ) GTPase that is involved in cytoskeletal dynamics and critically required for B cell development [42] ( Fig 2C and 2D ) . Additional noteworthy targets include a core NF-kappaB signaling component Ikk-beta ( IKBKB ) and members of the Mucosa-associated lymphoid tissue lymphoma translocation protein 1 ( MALT1 ) signaling complex ( MALT1 , BCL10 ) that modulate NF-kappaB activation ( Fig 2B ) . From the list of target genes related to BCR signaling , we selected several 3’UTRs for further investigation by conventional 3’UTR reporter assays; this included the GRB2 and MALT1 3’UTRs which are potentially regulated by multiple EBV miRNAs ( Fig 2A ) . With the exceptions of the CDKN1A and PRDM1 3’UTRs which are expressed from pLSG [18] , 3’UTRs are expressed from the dual luciferase reporter vector , psiCheck2 . Reporters were co-transfected with individual EBV miRNA expression vectors into 293T cells , lysates were collected 48–72 hrs post-transfection , and luciferase activity was measured . Luciferase knockdown was observed for 12 out of 15 PAR-CLIP-identified miRNA interactions , confirming these cellular 3’UTRs as targets of the EBV miRNAs ( Fig 3A–3C ) . To further investigate 3’UTR interactions for the BHRF1-2 miRNAs , we then used site-directed mutagenesis to disrupt individual seed-match sites , focusing our efforts on the MALT1 , GRB2 , RAC1 , INPP5D , and PAG1 3’UTRs . The Ago-CLIP identified interaction between miR-BHRF1-2-3p and the PRDM1 3’UTR was recently confirmed [17 , 43] . Disrupting the seed match sites fully alleviated miRNA-mediated luciferase knockdown in all instances ( Fig 3D ) thereby demonstrating that these are indeed bona fide interaction sites for the EBV BHRF1-2 miRNAs , and further specifying direct targeting by either the 3p or 5p miRNA ( both miRNAs are generated by the BHRF1-2 vector ) . Fig 3E shows examples of the mutations made for three of the 3’UTRs tested . Inhibition of the MALT1 3’UTR by EBV miR-BHRF1-2-5p in luciferase assays was recently reported [44]; here , our experiments confirm the PAR-CLIP studies that mapped the miRNA interaction site to a single site within the first 800 nt of the MALT1 3’UTR downstream of the stop-codon ( Fig 3E ) [17] . In total , we validated eight targets for the EBV BHRF1-2 miRNAs ( GRB2 , MALT1 , CDKN1A , PAG1 , INPP5D , RAC1 , PRDM1 , SOS1 ) that are associated with BCR signaling and further demonstrate through these assays that the MALT1 and CDKN1A 3’UTRs can be regulated by more than one EBV miRNA . To ask if the targets of EBV miRNAs validated above exhibited changes at the protein level , we first examined Grb2 protein levels in BHRF1-2 miRNA mutant LCLs or BJAB cells that were stably transduced with pLCE control vector or pLCE-BHRF1-2 to express the BHRF1-2 miRNAs in trans [11] . Consistent with inhibition of the GRB2 3’UTR luciferase reporter , we observed a ~25% decrease in Grb2 levels in LCLs and >70% reduction in Grb2 in BJAB cells upon introduction of the BHRF1-2 miRNAs ( Fig 3F and 3G ) . In BJAB cells , we also observed reduced phospholipase C gamma 1 ( Plcg1 ) levels , which we have recently demonstrated is regulated by miR-BHRF1-2-3p [11 , 17] . To test other targets , miRNA expression vectors were transfected into 293T cells and lysates were analyzed for protein expression . By Western blot assays , we observed a decrease in Malt1 levels ( Fig 3H ) in the presence of the BHRF1-2 miRNAs . These data , together with the luciferase reporter assays , thus provide experimental evidence for multiple targets of the EBV miRNAs that are commonly linked to BCR signaling pathways . To examine targets in the context of EBV infection , we also compared Grb2 and Malt1 protein levels in established , donor-matched EBV B95-8 wild-type LCLs versus BHRF1-2 miRNA mutant LCLs that lack both miR-BHRF1-2-3p and miR-BHRF1-2-5p [11 , 45] . Malt1 was upregulated in BHRF1-2 miRNA mutant LCLs for two out of three donor pairs and Grb2 was upregulated in BHRF1-2 miRNA mutant LCLs for one of the two donor pairs tested ( S1A and S1B Fig ) , demonstrating that loss of BHRF1-2 miRNA activity in latently infected cells can confer enhanced protein expression . We further assessed target RNAs by qRT-PCR in EBV B95-8 wild-type and BHRF1-2 miRNA mutant LCLs . For most targets , steady state levels were not reproducibly affected by the presence or absence of the BHRF1-2 miRNAs ( S1C–S1G Fig ) , suggesting that mRNA repression by the BHRF1-2 miRNAs may not generally lead to mRNA degradation . Alternatively , mRNA stability for these BHRF1-2 targets may be impacted during a specific stage of EBV infection that is not captured in established LCLs . We did , however , observe significant knockdown of GRB2 transcripts upon ectopic expression of the BHRF1-2 miRNAs in BJAB cells , demonstrating that GRB2 mRNAs may be subject to BHRF1-2 miRNA mediated degradation or decay under certain conditions ( S1H Fig ) . Having identified several BCR signaling components as EBV miRNA targets , we then asked whether RNAi-mediated knockdown of the individual target RNAs could potentially phenocopy EBV miRNA function . We generated shRNAs ( short hairpin RNAs ) for individual cellular targets of the EBV BHRF1-2 miRNAs since activity of both NF-kappaB and AP1 reporters was reduced by these miRNAs in response to anti-IgM stimulation ( Fig 1 ) . miR-30-based shRNAs generated against RAC1 , GRB2 , PLCG1 , SOS1 , INPP5D , and MALT1 were introduced into BJAB cells; qRT-PCR and/or Western blot analysis was performed to confirm that each shRNA construct appropriately knocked down their target of interest ( Fig 4A and S2A Fig ) . All shRNAs were then introduced into BJAB-NFkB-Luc reporter cells , lentiviral transduction was confirmed by mCherry fluorescence , and transcript levels were assayed by qRT-PCR to confirm knockdown ( Fig 4B ) . Cells were subsequently stimulated with anti-IgM to activate the NF-kappaB reporter . shMALT1 , shINPP5D , and shPLCG1 cells remained partially responsive to BCR cross-linking , while shRNA depletion of GRB2 , SOS1 , or RAC1 potently inhibited NF-kappaB activation ( Fig 4C ) . Notably , GRB2 , SOS1 , and RAC1 are all targets of miR-BHRF1-2-5p ( Fig 3 ) , suggesting that this miRNA , in particular , acts as a key regulator of BCR signal transduction . Previous studies show that genetic ablation of the EBV BHRF1 miRNAs from the viral genome impairs but does not fully inhibit LCL outgrowth in vitro [45–48] . Furthermore , LCLs established with BHRF1 miRNA mutant viruses exhibit altered cell cycle progression and reduced growth rates [46 , 47] . While these phenotypes can be attributed , in part , to inefficient processing of the BHRF1 RNAs and/or deregulated splicing of EBNA-LP transcripts that occurs in cis [49 , 50] , the contributions from specific cellular targets of the BHRF1 miRNAs have not been fully examined . GRB2 encodes a ubiquitously expressed signaling adaptor protein that is recruited to growth factor receptors via its SH2 and SH3 domains , coordinates with SOS1 to recruit Ras GTPases , and subsequently , transduces external signals that can activate genes involved in cell proliferation [40 , 41] . We therefore hypothesized that targets of the BHRF1-2 miRNAs , such as GRB2 or SOS1 , might play a role in the survival and/or proliferation of latently infected B cells . Using cell viability assays ( trypan-blue exclusion ) as well as MTT ( 3- ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ) assays to monitor metabolic activity , we first measured the growth of established , donor-matched LCLs ( >6–8 weeks post-infection ) that are latently infected with either EBV B95-8 2089 wild-type virus or a miRNA mutant virus in which both BHRF1-2 miRNAs are deleted [45] . In agreement with prior studies [45 , 46] , LCLs established with the BHRF1-2 miRNA mutant virus grew significantly slower than wild-type LCLs ( S3A–S3C Fig ) . We subsequently asked whether this growth phenotype could be rescued by re-introducing the BHRF1-2 miRNAs . We measured the growth rates of two BHRF1-2 miRNA mutant LCLs which stably express BHRF1-2 miRNAs in trans [11] . Compared to control cells transduced with pLCE empty vector , ectopic BHRF1-2 miRNA expression enhanced LCL growth rates by 20–25% ( Fig 5A and S3D and S3E Fig ) . In an effort to improve the differences in cell proliferation observed between control and BHRF1-2 miRNA expressing cells , we also tested growth rates in 20% serum conditions . As anticipated , LCLs exhibited increased growth rates at higher serum concentrations; however , the measured effect of the BHRF1-2 miRNAs was comparable to the 10% serum conditions , enhancing cellular growth rates by ~20% ( Fig 5A ) . BHRF1-2 miRNA expression in EBV negative BJAB cells had no significant effect on cell growth ( S3F Fig ) , suggesting this phenotype may be specific to the B cell stage of LCLs . These results show that the growth defect conferred through genetic loss of BHRF1-2 miRNAs from the viral genome can be at least partially reversed by complementing mutant LCLs with the BHRF1-2 miRNAs in trans . To examine whether cellular factors identified above might contribute to LCL proliferation , we then introduced shRNAs into BHRF1-2 miRNA mutant LCLs ( LCL-D2 ) against GRB2 , MALT1 , INPP5D , SOS1 , RAC1 , or IL1R1 ( regulated by miR-BHRF1-2-5p [11] ) . Following shRNA transduction , MTT and cell viability assays were performed to monitor changes in growth patterns ( Fig 5B–5D ) . Knockdown of each target transcript was confirmed by qRT-PCR analysis ( S2B Fig ) . MALT1 inhibition had no observable effects on LCL growth , despite achieving ~50% knockdown with this shRNA ( S2B Fig ) . Strikingly , shRNA inhibition of INPP5D , SOS1 , IL1R1 , or RAC1 in LCL-D2 was detrimental to cell growth , while in three separate experiments , shRNA-mediated reduction of GRB2 lead to a small , but statistically significant increase in LCL-D2 growth ( Fig 5B–5D ) . These results demonstrate dependencies on INPP5D , SOS1 , IL1R1 , and RAC1 for efficient LCL proliferation and furthermore , indicate that RNAi-mediated control of GRB2 confers a modest growth advantage to latently infected B cells . The GRB2 3’UTR is targeted by miR-BHRF1-2-5p in luciferase reporter assays ( Fig 3 ) . To confirm that this miRNA specifically controls Grb2 expression during EBV infection , we performed Western blot assays . Grb2 protein levels were increased by >2-fold in LCLs transduced with a sponge inhibitor for miR-BHRF1-2-5p ( Fig 5E ) . To examine the relationship between miR-BHRF1-2-5p and LCL growth , we then introduced the miR-BHRF1-2-5p inhibitor into additional LCLs infected with either EBV B95-8 or a wild-type EBV . miRNA knockdown was confirmed by Taqman qRT-PCR and importantly , levels of miR-BHRF1-2-3p were unaffected ( Fig 5F ) . MTT assays were subsequently performed to monitor metabolic activity of miRNA-sponged LCLs . Loss of miR-BHRF1-2-5p function did not critically impair LCL growth , such has previously been observed for miR-155 loss [51]; however , inhibition of the miRNA did significantly attenuate the growth of all three LCLs ( Fig 5G ) . In separate experiments , we monitored viable cell counts for miRNA-sponged LCLs and controls ( pLCE-CXCR4s or pLCE empty vector ) generated in parallel ( S4A–S4C Fig ) . Congruent with MTT assay results , inhibition of miR-BHRF1-2-5p activity reproducibly reduced cell growth rates by 20–30% ( S4A–S4C , S4H and S4I Fig ) , demonstrating that latently infected B cells are dependent on miR-BHRF1-2-5p function for efficient proliferation . To ask if we could rescue the growth phenotype conferred by miR-BHRF1-2-5p knockdown , GRB2 shRNAs were expressed in combination with the miR-BHRF1-2-5p sponge and MTT assays were performed ( Fig 5G ) . Consistent with the phenotype observed in BHRF1-2 miRNA mutant LCLs ( Fig 5B ) , shRNA-mediated inhibition of GRB2 reversed the miRNA inhibitor effects and led to increased proliferation of miR-BHRF1-2-5p-sponged LCLs compared to control cells ( Fig 5G and S4D–S4G Fig ) . The increase in cell proliferation was specific to GRB2 since shRNAs against RAC1 or SOS1 did not have a measurable impact on LCL growth ( S4D and S4E Fig ) . Together , these results demonstrate that EBV miR-BHRF1-2-5p plays an active role in promoting the growth of latently infected B cells through the regulation of GRB2 . As LCLs are artificially generated in vitro , we next sought to determine whether miR-BHRF1-2-5p also plays a role in maintaining the growth of EBV transformed B cells that are naturally derived . Diffuse large B cell lymphoma ( DLBCL ) is the most common form of NHL in HIV-infected patients , and >85% of immunoblastic , AIDS-related DLBCL are EBV+ [52] . Using qRT-PCR , we first characterized the viral gene expression pattern in EBV+ DLBCL cell lines [53] . Consistent with a latency III program , both the LMP1 and EBNA2 transcripts were present , as well as high levels of the BHRF1 miRNAs ( S5 Fig ) . We then transduced IBL1 and BCKN1 cells with the miR-BHRF1-2-5p sponge inhibitor . Transduced DLBCL cells expressing high GFP were FAC-sorted , and miRNA knockdown was confirmed by qRT-PCR ( Fig 6A ) . Cell proliferation was subsequently measured by MTT assay ( Fig 6B ) . Similar to LCLs , sponge inhibition of miR-BHRF1-2-5p in the DLBCL cell lines attenuated cell proliferation ( Fig 6B ) . To corroborate these results , we also performed growth curves with FAC-sorted BCKN1 cells , and observed that loss of miR-BHRF1-2-5p activity reduced cell growth by ~20% compared to control cells ( Fig 6C ) . To examine the possibility that spontaneous lytic reactivation was occurring in response to miR-BHRF1-2-5p inhibition and contributing to the decrease in cell proliferation , we monitored viral gene expression in both DLBCLs and LCLs . Neither BZLF1 nor EBNA2 levels were altered in the sponged cells ( Fig 6D and 6E ) , ruling out virus reactivation as a contributing factor in the regulation of B cell proliferation by EBV miR-BHRF1-2-5p . Cross-linking of surface Ig on B cells latently infected with EBV can facilitate lytic reactivation [54] . To investigate miRNA activities in EBV+ B cells that are sensitive to BCR stimuli , we first examined miRNA expression in Mutu I BL cells which , like BJAB cells , express membrane-bound IgM on the surface but are latently infected with EBV type 1 and characteristically exhibit a restricted latency type I pattern of viral gene expression ( i . e . EBNA1 , low levels of BART miRNAs , nearly undetectable levels of BHRF1 miRNAs ) [8 , 55] . Treatment of Mutu I cells with anti-IgM resulted in ~25–35 fold increases in EBV immediate-early , early , and late gene products ( BZLF1 , BRLF1 , BALF4 , BNLF2a , and BHRF1 ) as detected by qRT-PCR ( S6A Fig ) , demonstrating entry into the lytic phase . Consistent with previous reports , we observed rapid induction of miR-BHRF1-2-5p , which correlated with BHRF1 mRNA levels , and a delay in the induction of miR-BHRF1-1-5p following lytic reactivation [8 , 56] ( S6B Fig ) . Notably , all EBV BART miRNAs that functionally inhibited BCR signaling ( miR-BART2-5p , miR-BART9 , miR-BART17 , and miR-BART18 ) were upregulated during lytic reactivation ( S6B and S6C Fig ) . We then tested whether EBV miRNAs could directly influence BCR-mediated EBV reactivation from latency , using sponge inhibitors to disrupt the activity of individual miRNAs in Mutu I cells . We focused on two EBV miRNAs: ( i ) miR-BHRF1-2-5p , for which we were able to confirm multiple targets directly linked to BCR signaling and ( ii ) miR-BART2-5p , which reduces BCR-mediated NF-kappaB activation ( Fig 1 ) and also targets EBV BALF5 [57] . Additionally , we included inhibitors for miR-BHRF1-1 , which was previously reported to enhance virus reactivation by targeting the ubiquitin ligase RNF4 [58] , and two cellular miRNAs- namely , the miR-17/20 family that regulates EBV BHRF1 and LMP1 [17 , 19 , 32] and miR-190 that was reported to inhibit phorbol ester or BCR-mediated reactivation of EBV+ BL cells [59] . An inhibitor to cellular miR-19 was included as an additional control . Mutu I cells were transduced with pLCE control or the individual miRNA sponge inhibitors , then stimulated with anti-IgM for 24 hr or 42 hr ( Fig 7A ) . We observed ~40–50% reduction in steady-state , mature miRNA levels in the presence of the sponge inhibitors , indicating the sponges were functional and in this experimental context , contributing to miRNA decay ( Fig 7B ) . To monitor lytic reactivation in MutuI cells , BZLF1 and BRLF1 transcripts were measured by qRT-PCR . Surface Ig cross-linking led to ~8 to 10-fold induction of EBV IE gene expression in pLCE control cells by 24 hr ( Fig 7A and 7C ) . We observed comparable ~10-fold induction of EBV IE transcripts in the presence of miR-BHRF1-1 , miR-BART15-3p , or miR-19 inhibitors , as well as the miR-190 inhibitor ( contrary to a previous report [59] ) , demonstrating these miRNAs do not significantly impact the initial stages of EBV reactivation in Mutu I cells . Upon suppression of miR-BHRF1-2-5p , miR-BART2-5p , or miR-17/20 activity , however , we observed significant increases in BZLF1 and BRLF1 levels following anti-IgM treatment compared to control cells ( Fig 7A and 7C ) . To investigate this further and ensure that the full lytic program was initiated , we measured additional early and late viral transcripts ( BALF4 , BNLF2a , EBNA2 , LMP1 ) in anti-IgM treated cells . Compared to control cells , all four EBV transcripts were significantly upregulated at both 24 hr and 42 hr in cells with inhibitors to miR-BHRF1-2-5p , miR-BART2-5p , or miR-17/20 ( Fig 7D–7G ) . These results demonstrate that miR-BHRF1-2-5p , miR-BART2-5p , and miR-17/20 regulate the amplitude of virus reactivation as triggered through surface Ig cross-linking . The overall increase in lytic reactivation upon miR-BART2-5p inhibition could be attributed , in part , to the targeting of EBV BALF5 encoding the viral DNA polymerase catalytic subunit [57] . To separate this activity from effects of miR-BART2-5p on the BCR signaling pathway , we also measured lytic reactivation at time points earlier than 24 hr . Increased BZLF1 levels were detectable in miR-BART2-5p sponged cells as early as 5 hours post anti-IgM stimulation , which occurs prior to BALF5 expression and virus replication , suggesting that miR-BART2-5p is indeed hindering reactivation signals mediated through BCR signals ( S7 Fig ) . Intriguingly , we did not see significant changes in basal BZLF1 or BRLF1 expression in mock treated Mutu I cells for any of the miRNA inhibitors ( Fig 7A and 7C ) . These data lead us to conclude that EBV miRNAs such as miR-BHRF1-2-5p and miR-BART2-5p are not actively blocking spontaneous lytic replication , but are positioned to respond to and antagonize extracellular stimuli that lead to virus reactivation ( Fig 8 ) .
In this study , we evaluated the role of EBV miRNAs in BCR signaling , which has implications for the proliferation of latently infected B cells and importantly , for influencing virus reactivation from the latent state . Antigen stimulation of the BCR induces multiple intracellular signaling pathways , and latent viral proteins have been previously demonstrated to manipulate BCR signaling components and/or the pathways ( i . e . NF-kappaB , AP1 , PI3K , MAPK ) triggered through BCR activation . As viral miRNAs are non-immunogenic viral gene products that are expressed throughout multiple stages of EBV infection , we hypothesized that the viral miRNAs would be prime candidates to coordinately modulate the signaling cascade initiated through the BCR . Through functional screens , we demonstrate that a subset of five EBV miRNAs desensitizes BL cells to BCR cross-linking , subsequently attenuating downstream transcriptional activation of NF-kappaB and/or AP1 ( Fig 1 ) . By analyzing existing B cell miRNA targetome datasets which globally captured EBV miRNA interactions [17 , 18 , 60] , we find that signal transducers as well as multiple components of the pathways situated downstream of the BCR can be regulated by the viral miRNAs . 54 host targets were identified as being associated with BCR pathways ( Figs 2 and 3 ) , and using biochemical assays , we validated several new interactions between EBV miRNAs and the cellular 3’UTRs ( GRB2 , PAG1 , RAC1 , INPP5D , IKBKB , CDKN1A , FOXO3 ) ( Fig 3 ) , thereby providing insight into the underlying molecular mechanisms by which these miRNAs attenuate BCR signaling . Our experiments highlight that the EBV BHRF1-2 miRNAs , in particular , act as novel and potent regulators of BCR signal transduction and consequently , can restrict entry into the lytic cycle initiated by BCR engagement . The BHRF1-2 miRNAs are evolutionarily conserved , with homologs encoded by multiple lymphocryptoviruses ( LCV ) that infect Old World non-human primates [32 , 61] . The high sequence conservation with other LCV BHRF1-2 miRNA homologs supports the notion that these miRNAs are of particular importance to the viral life cycle . To date , there are only a few characterized targets of the EBV BHRF1-2 miRNAs and even fewer targets linked to function [11 , 17 , 43 , 62] . Here , we evaluated several interactions between the BHRF1-2 miRNAs and host 3’UTRs ( MALT1 , PRDM1 , SOS1 , PLCG1 ) that have been examined in other reports in the literature [11 , 43 , 44] , and further confirmed novel interactions with the GRB2 , RAC1 , PAG1 , and INPP5D 3’UTRs ( Fig 3 ) . Site-directed mutagenesis of miRNA seed-match sites in several of these 3’UTRs abrogated luciferase reporter knockdown , clearly demonstrating that the BHRF1-2 miRNAs interact with these target RNAs through the identified binding sites . Commonly , Grb2 , Plcg1 , Sos1 , and Pag1 interface with receptor tyrosine kinases ( RTKs ) to link the BCR with downstream pathways . The fact that multiple players in BCR responses are inhibited by the BHRF1-2 miRNAs can thus explain the strong disruption of BCR signaling . Grb2 is a ubiquitously expressed signaling adaptor protein that interacts with numerous growth factor and antigen RTKs via its src-homology domains [40 , 63] . The role of Grb2 in regulating BCR signaling and normal B cell responses ( i . e . not in the context of EBV infection ) is not completely understood . Grb2 was initially described as a negative regulator of B cell activation [63] , following studies demonstrating that ablation of Grb2 from mature B cells in mice enhanced proliferation responses to BCR cross-linking and altered lymphoid follicle organization in germinal centers [64 , 65] . Intriguingly , B cell-specific Grb2-/- mice exhibited reduced numbers of splenic B cells in the periphery as well as reduced follicular B cells , and lacked germinal centers in the spleen but not in other secondary lymphoid organs [40 , 64] . More recent studies demonstrate that Grb2 acts as a positive regulator of B cell activation; re-introduction of the protein into Grb2-deficient splenic murine B cells enhanced BCR-induced calcium mobilization [66] . In reconciling these observations , it is likely that Grb2 has a differential role in B cell activation where cell context and stage of B cell development are critical factors . Studies on FGFR2 ( fibroblast growth factor receptor 2 ) expressing cancer cells demonstrate that Grb2 function is concentration dependent and when interacting with RTKs , focally concentrated Grb2 levels contribute to receptor pre-dimerization in the absence of external stimulation while at the same time prevent uninitiated downstream responses [67] . Of note , Plcg1 has been shown to compete with Grb2 for RTK interactions; in the context of FGFR2 signaling , knocking down Grb2 allows for increased recruitment of Plcg1 to the receptor which has implications for increased metastatic potential [67] . In human BL cells , we observed that shRNA-mediated knockdown of GRB2 , but not PLCG1 , inhibited BCR-triggered signaling events , thereby phenocopying BHRF1-2 miRNA activity ( Fig 4 ) . Thus , in this context , our findings are in line with Grb2 acting as a positive regulator to amplify BCR-induced signals . Notably , GRB2 as well as SOS1 are regulated by non-coding RNAs in other g-herpesvirus infection systems , suggesting this may be a common strategy employed by g-herpesviruses to manipulate RTK signaling . KSHV miR-K10a , a mimic of cellular miR-142-3p , targets the SOS1 3’UTR [18] . KSHV miR-K12-4-3p and miR-K9-5p target the GRB2 3’UTR and reduce Grb2 protein levels during KSHV infection [18 , 68] . In contrast , Herpesvirus saimiri ( HVS ) , a non-human primate g-herpesvirus that naturally infects squirrel monkeys and is transforming in marmoset T cells , upregulates Grb2 levels by counteracting the activity of cellular miR-27 [69] . While GRB2 has been shown to be regulated by these other g-herpesviruses [18 , 68 , 69] , the functional significance of these interactions has not been fully determined . Grb2 is conventionally linked to Ras and MAPK signal transduction pathways through its cytoplasmic binding partners , such as PLCG1 and the Ras guanine nucleotide exchange factor encoded by SOS1 . In our experiments , we found that shRNAs knocking down GRB2 or SOS1 in particular phenocopied BHRF1-2 miRNA function and had a negative impact on NF-kB activation in response to BCR cross-linking . While this leads us to conclude that disruption of GRB2 or SOS1 by RNA-interference mechanisms impairs antigen receptor signaling , the fact that NF-kappaB activity was reduced is not so easily explained . Due to the numerous binding partners of Grb2 , we speculate one way this might occur is through indirect effects of secondary messengers , such as DAG ( diacylglycerol ) and IP3 ( inositol-1 , 4 , 5-triphosphate ) , that are generated from PIP2 ( phosphatiyl-4 , 5-bisphosphate ) catalysis in response to BCR triggering . DAG activates protein kinase C which can subsequently activate NF-kappaB , among other factors . Thus , apart from the canonical Ras and MAPK pathways , impairment of a BCR signal integrator such as Grb2 has multiple apparent consequences for downstream signaling pathways . Through loss-of-function experiments , our study further reveals that miR-BHRF1-2-5p activity and optimal Grb2 levels are necessary for efficient proliferation of EBV transformed B cells . In concordance with previous reports , LCLs established with BHRF1-2 miRNA mutant viruses exhibited growth defects compared to those established with wild-type EBV B95-8 [45 , 48] ( Fig 5 and S3 Fig ) . While these previous reports have speculated that inhibition of cellular factors by the BHRF1 miRNAs is necessary to provide a favorable environment for effective B cell transformation and/or growth of LCLs , the cellular targets have been elusive . Work from the Delecluse lab demonstrated that mutation of the region encoding the BHRF1-2 miRNAs affects processing of the adjacent BHRF1-3 primary miRNA; thus , B cells infected with BHRF1-2 miRNA mutant viruses exhibit reduced levels of miR-BHRF1-3 when compared to B cells infected with wt B95-8 [48] . This complicates interpretation of experiments measuring B cell transformation efficiencies at early stage infection when viruses harboring mutations in individual or multiple BHRF1 miRNAs are compared [45 , 48 , 49] . By comparing different iterations of BHRF1 miRNA mutant viruses , these prior studies proposed that miR-BHRF1-3 is a vital part of EBV-mediated B cell transformation [48 , 49] . Our results argue that the conserved BHRF1-2 miRNAs are the primary contributors in maintaining the growth of transformed B cells ( Figs 5 and 6 , S4H Fig ) . We tested whether miR-BHRF1-3 could be a contributing factor at later stages by measuring the growth of LCLs generated with BHRF1-2 , BHRF1-3 , or triple BHRF1 miRNA mutant viruses ( S3 Fig ) . Consistent with previous studies , LCLs lacking only the BHRF1-2 miRNAs or all three BHRF1 miRNAs had dramatically reduced growth capabilities compared to wt virus; however , we did not observe any growth differences with the BHRF1-3 miRNA mutant virus . This discrepancy could be explained by the time points examined; it is possible that growth defects associated with abrogation of miR-BHRF1-3 occur at early stages after de novo infection and may resolve later once LCLs become established . Recent work from Poling et . al . investigated whether growth deficiencies in LCLs generated with a virus lacking all BHRF1 pre-miRNAs could be reversed [50] . In contrast to our experiments which focused specifically on a mutant lacking only the BHRF1-2 miRNAs , simultaneous expression of all BHRF1 miRNAs in trans failed to rescue the triple mutant . Molecular analysis by Poling et . al . further revealed that mutations in the BHRF1 pre-miRNA stem loops induce altered splicing patterns in the EBNA-LP transcripts , which the authors rationalized were irreversibly detrimental to LCL growth [50] . Of note , EBNA-LP levels are also upregulated in viral mutants where the BHRF1-2 or BHRF1-3 miRNAs are individually inactivated [45] . We found that LCL growth was crippled predominantly when the BHRF1-2 miRNAs were absent , but not the BHRF1-3 miRNAs ( Fig 5 and S3 Fig ) , presenting the likelihood that factors in addition to EBNA-LP are responsible for changes in growth patterns . We used multiple strategies , including shRNAs as well as miR-BHRF1-2-5p sponge inhibitors , to demonstrate that miRNA-mediated control of host targets such as GRB2 indeed provides a growth advantage for latently infected B cells . Long-lived , memory B cells are thought to be a primary reservoir for persistent EBV infection in vivo as the virus is able to gain access to this compartment through viral gene expression programs that manipulate B cell growth and differentiation [5] . While the stimulus for EBV reactivation in vivo is not fully known , activation of BCR signaling is thought to represent the most physiologically relevant trigger for lytic reactivation in vitro . In examining the functional consequences of perturbing miRNA activity , we uncovered roles for EBV miR-BHRF1-2-5p , miR-BART2-5p , and unexpectedly , miR-17-5p in the latent to lytic switch ( Fig 7 ) . Herpesviruses have complex relationships with miR-17 family members . Human cytomegalovirus , for example , induces selective turnover of miR-17 family members via a viral decay element in order to accelerate lytic infection [70] . Our data show that inhibition of miR-17 can also accelerate lytic replication for EBV . In contrast to CMV , EBV and other g-herpesviruses do not appear to have mechanisms for inducing miR-17 turnover , and in fact , seem to require miR-17 or miR-17-like activity during latent infection [71 , 72] . Furthermore , EBV miRNAs collectively share a significant number of targets with miR-17 [19] , arguing that repression of miR-17 target genes is a key part of the latent g-herpesvirus life cycle . Although the mechanism ( s ) by which miR-17 controls EBV reactivation remain to be elucidated , we note that Ago-CLIP studies have cataloged miR-17 targets such as SOS1 , VAV2 , and GAB1 [17–19] that indicate the phenotype , like that of the EBV miRNAs , is linked to cellular responses associated with BCR triggering . By suppressing signals that arise via engagement of the BCR , we propose that EBV miRNAs such as miR-BHRF1-2-5p and miR-BART2-5p protect latent cells from aberrant reactivation ( Fig 8 ) . We further postulate that the BHRF1-2 miRNAs play additional roles in dampening lytic reactivation . Cross-linking of the BCR activates latently infected memory B cells to differentiate into plasma cells [23] . Previous studies report that miR-BHRF1-2-3p targets the PRDM1 3’UTR , which we confirmed in this study using luciferase reporter assays [17 , 43] ( Fig 3 ) . Blimp1 , encoded by PRDM1 , is a master B cell transcription factor required for the differentiation of B cells into plasma cells . Overexpression of Blimp1 in several BL cell lines , particularly Wp-restricted BL , is sufficient to induce lytic reactivation , which is thought to occur through Blimp1-mediated activation of the IE promoters driving BZLF1 ( Zp ) and BRLF1 expression [24] . While it remains to be tested , miR-BHRF1-2-3p could potentially suppress lytic reactivation by repressing Blimp1 . It is conceivable that 5p miRNA , miR-BHRF1-2-5p , also attenuates EBV lytic replication by regulating transcription factors involved in B cell differentiation . SP1 ( specificity protein 1 ) , for example , is a cellular transcription factor that also binds to and activates Zp [73] , and a binding site for miR-BHRF1-2-5p was previously reported in the SP1 3’UTR [17] . Given that we did not observe spontaneous lytic reactivation upon miR-BHRF1-2-5p inhibition in LCLs , DLBCLs , or BL ( Figs 6 and 7 ) , we speculate that cellular context as well as cues from the extracellular environment are major participants in miRNA-influenced decisions impacting virus reactivation . An intriguing question that remains is how miR-BHRF1-2-5p physically exerts function in latency I BL cells during induction of the lytic cycle . Unlike miR-BART2-5p and cellular miR-17 , the BHRF1-2 miRNAs are not detectable in latency I until cells respond to reactivation cues ( S6B Fig ) ; yet , introduction of an inhibitor against miR-BHRF1-2-5p enhances lytic gene expression ( Fig 7 ) . One possible explanation is that within this heterogeneous population , the response of individual cells to reactivation stimuli is asynchronous . In this scenario , some cells will conceivably respond more quickly and express the BHRF1-2 miRNAs before other cells become responsive . Previous reports have demonstrated that the EBV BHRF1 miRNAs can be transferred between EBV+ B cells and uninfected T cells in co-culture experiments [74] . Moreover , delivery of functional EBV miRNAs via exosomes has been demonstrated by several groups [75 , 76] . While further investigation will be required , we speculate that the BHRF1-2 miRNAs may be transferred to neighboring cells to exert their effects , thereby controlling the overall level of virus reactivation in the cell pool . In summary , we have shown that a subset of EBV miRNAs attenuates signaling through the BCR and governs aspects of EBV reactivation . An important goal for future studies will be to examine specific contributions from additional viral miRNAs as well as cellular miRNAs that are hijacked by the virus in order to more fully decipher the extent to which post-transcriptional regulatory mechanisms facilitate latent/lytic cell state transitions . Gaining a clear understanding of the mechanisms regulating latency and entry into the lytic replication cycle is a critical part of designing effective treatments for viral disease . Our findings thus provide valuable insight into how viral miRNAs functionally contribute to viral latency and have utility in aiding development of miRNA-focused therapeutic strategies .
B cell lines ( BJAB , LCLs , DLBCLs , MutuI ) were maintained at 37°C in a 5% CO2-humidified atmosphere in RPMI-1640 supplemented with 15% fetal bovine serum ( FBS ) ( unless otherwise stated ) and 1% penicillin , streptomycin , and L-glutamine ( P/S/G ) . MutuI BL cells exhibit cell surface IgM ( uK+ ) [55] and were kindly provided by Dr . Erik Flemington at Tulane University . BJAB ( EBV-neg germinal-center derived BL cell line ) , EBV wild-type , EBV B95-8 and BHRF1-2 miRNA mutant LCLs ( IBL-LCLd3 , SDLCL , LCL35 , LCLBACWT , LCLBACD2 ) originated from the laboratories of Dr . Bryan Cullen or Dr . Micah Luftig at Duke University and are previously described in [17 , 32 , 60] . Additional LCLs were generated with EBV B95-8 2089 or EBV BHRF1 miRNA mutant viruses , with kind permission from Dr . H . J . DeLecluse at the German Cancer Research Centre , using a multiplicity of infection = 2 Raji-GFP units [11 , 45]; LCL-WT , LCL-D2 , LCLWT-16 . 1 , and LCLD2-16 . 1 are previously described in [11] . DLBCL cell lines ( IBL1 , BCKN1 , IBL4 ) are previously described [53] and were kindly provided by Dr . Ethel Cesarman at Weill Medical College of Cornell University . Human primary B lymphocytes ( PBMCs ) used in this study were isolated from anonymous whole blood purchased from Research Blood Components . HEK293T cells ( originating from Dr . Bryan Cullen’s laboratory ) were maintained in high glucose DMEM supplemented with 10% FBS and 1% P/S/G . For preparation of lentiviruses , HEK293T cells were plated in 15-cm plates in complete media and transfected using Polyethylenimine ( PEI ) with 7 . 5 ug pL-based lentivector , 4 . 5 ug pDeltaR8 . 75 and 3 ug pMD2G . Media was changed to complete RPMI-1640 between 8 hrs and 16 hrs post-transfection . Lentiviral particles were harvested by sterile filtration of the supernatant using a 0 . 45 micron filter at 48 and 96 hrs post-transfection and used to transduce ~1 to 5 x 10^6 cells . BJAB , MutuI , LCLs , and DLBCLs were transduced with one ( BJAB ) or two ( LCLs , DLBCLs , and MutuI ) rounds of lentivirus and monitored by fluorescent microscopy for transduction efficiency by green fluorescent protein ( GFP ) and/or mCherry expression . DLBCLs were further subjected to sorting by flow cytometry to obtain >95% pure populations of GFP-positive cells . To induce lytic reactivation , MutuI cells were spun down and plated at 1 x 10^6 cells in fresh media containing soluble anti-IgM ( Sigma ) at concentrations and times indicated in figure legends ( 2 . 5–5 ug/mL for 22–48 hrs ) . EBV miRNA expression vectors in pLCE are previously described [16 , 17 , 32] . pLSG-PRDM1 is previously described [18] and was generously provided by Dr . Eva Gottwein at Northwestern University . psiCheck2-RAC1 and psiCheck2-IKBKB are previously described [77 , 78]; psiCheck2-FOXO3 . 2 ( containing the second half of the FOXO3 3’UTR ) was generously provided by Dr . Jay Nelson’s laboratory at Oregon Health and Science University . To generate other 3’UTR luciferase reporters , 3’UTRs were PCR amplified from genomic DNA of EBV-infected B cells and cloned into the XhoI and NotI sites downstream of Renilla luciferase in the psiCheck2 dual luciferase reporter vector . When achievable , the entire 3’UTR was cloned for a given target . For longer 3’UTRs , a minimum of 1 kb containing the region predicted to be targeted by each miRNA was cloned . Mutant 3’UTR reporters , containing nucleotide changes in miRNA seed match sites as identified by PAR-CLIP , were generated by Phusion Taq site-directed mutagenesis . Lentiviral vector miRNA sponge inhibitors contain six to eight imperfect , tandem decoy binding sites for a single miRNA as previously described [11 , 17 , 79] . Sponge oligonucleotides contain flanking KflI sites and were annealed and concatamerized prior to cloning into the GFP 3’UTR of pLCE . The control sponge ( pLCE-CXCR4s ) is previously described [17 , 79] . miR-30-based shRNA constructs were cloned into the XhoI/EcoRI sites of pL-CMV-mCherry vector using GRB2 . 3140 , MALT1 . 3197 , RAC1 . 424 , INPP5D . 1191 oligonucleotides as described in [80] . shRNA constructs for PLCG1 and SOS1 are previously described [11] . Cells were maintained in log phase and split 1:2 or 1:3 one day prior to seeding into 96-well plates at 2 . 5 x 10^3 cells/well in complete RPMI media without phenol red . At each time point , 100 ul per well of MTT ( 3- ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazoliumbromide ) reagent ( 5 mg/ml in phosphate-buffered saline ) was added . Cells were incubated for 2 hrs at 37°C , lysed in MTT solvent ( isopropanol containing 0 . 5% NP40 and 4 mM HCl ) , and incubated for an additional 2 hrs at 37°C . Lysates were read using an ELISA plate reader ( Absorbance = 562 nm ) , and relative cell growth was determined by comparing absorbance values of cell lysates at time zero ( T0 ) to the values at 24 , 48 , or 72 hrs post-plating ( Tn ) as indicated . Growth curves were performed 4 to 7 days after final transductions when LCLs were >75% GFP positive by microscopy and in the log-phase of growth . FAC-sorted DLBCLs were rested for 2 days after sorting and prior to plating for growth curves , To assay growth , cells were split 1:2 in BJAB-conditioned complete media one day prior to plating and plated in triplicate or quadruplicate at 2 x 10^5 cells per mL in 24-well plates . Viable cells were counted by trypan blue exclusion using a hemacytometer at time points indicated . Cells were lysed in NP40-lysis buffer ( 50 mM HEPES pH 7 . 5 , 150 mM KCl , 2 mM EDTA , 1 mM NaF , 0 . 5% ( vol/vol ) NP40 , 0 . 5 mM dithiothreitol ( DTT ) ) and protein concentrations determined using the BCA protein assay kit ( Thermo Scientific ) . 20 ug of total protein per lane was resolved on 10% Tris-glycine SDS-PAGE and transferred onto Immobilon PVDF membranes . Following blocking in 5% milk in TBS-Tween , blots were probed with primary antibodies to Grb2 ( #3972 , Cell Signaling Technology ) , Malt1 ( B-12 , sc-46677 , Santa Cruz ) , Plcg1 ( D9H10 , #5690 , Cell Signaling Technology ) , beta-Actin ( clone C4 , sc-47778 , Santa Cruz ) , or Gapdh ( #ab8245 , Abcam ) , then probed with horse-radish peroxidase ( HRP ) conjugated secondary antibodies ( anti-rabbit IgG or anti-mouse IgG ) . Blots were developed using enhanced chemiluminescent substrate ( Pierce ) and exposed to film . Protein levels were quantified from scanned films using NIH ImageJ . NF-kappaB and AP1 activity was assayed using BJAB luciferase reporter cell lines . BJAB-NFkBLuc cells are previously described [11] . To generate BJAB-NFkB-GL4 . 32 or BJAB-AP1-GL4 . 44 , 3 ug linearized plasmid ( pGL4 . 32 or pGL4 . 44 ( Promega ) ) was transfected into BJAB cells using the Amaxa Nucleofector II device ( Lonza ) , program G-016 , in 100 ul of Ingenio electroporation solution ( Mirus ) . Cells were placed under hygromycin selection for two weeks . EBV miRNAs or shRNAs were subsequently introduced by lentiviral transduction . For NF-kappaB activation , 1 x 10^5 cells per well were plated in 96-well black-well plates , stimulated for 18 hrs with 5 or 10 ug/ml anti-IgM ( Sigma ) as indicated , and then lysed in 1X passive lysis buffer ( Promega ) . Luciferase activity was assayed using the Dual Luciferase Reporter Assay System ( Promega ) and a Veritas microplate luminometer ( Turner Biosystems ) with dual injectors . For AP1 activation , cells were stimulated for 6 hrs with 10 ug/ml anti-IgM ( Sigma ) , and lysed in 1X passive lysis buffer ( Promega ) . AP1 firefly luciferase activity is normalized to total protein levels quantified with the BCA Protein Assay kit ( Pierce ) . HEK293T cells plated in 96-well black-well plates were co-transfected with 20 ng of 3’UTR reporter and 250 ng of control vector ( pLCE ) or EBV miRNA expression vector [16 , 17 , 32] using Lipofectamine2000 ( Thermofisher ) according to the manufacturer’s instructions . 48–72 hrs post-transfection , cells were collected in 1X passive lysis buffer ( Promega ) . Lysates were assayed for luciferase activity using the Dual Luciferase Reporter Assay System ( Promega ) and a luminometer with dual injectors . All values are reported as relative light units ( RLU ) relative to luciferase internal control and normalized to pLCE control vector . Total RNA was extracted using TRIzol ( Thermofisher ) . For detection of cellular or EBV transcripts , RNA was DNAse-treated and reversed transcribed using MultiScribe ( Thermofisher ) with random hexamers . Transcripts were detected using SYBR Green qPCR and oligonucleotides designed to amplify gene specific regions of ~200 bp . Primers for amplification of EBV transcripts BZLF1 , BRLF1 , BALF4 , BNLF2a , and BHRF1 are described in [81] . Oligonucleotides are listed in Table S1 . miRNAs were detected using commercial Taqman assays or miRNA stem-loop qRT-PCR assays as previously described [11 , 49] . miRNA levels are reported relative to miR-16 ( assay #000391 , Thermofisher ) or U6 ( assay #001973 , Thermofisher ) as indicated . All PCR reactions were performed in technical replicates ( duplicates or triplicates ) . PAR-CLIP datasets for EBV+ BC1 cells and EBV B95-8 or wild-type LCLs are previously described [17 , 18 , 32 , 60] . Raw fastq files were preprocessed as described previously and reads ≥13 nt were aligned to the human genome ( HG19 ) and either the EBV B95-8 genome ( V01555 . 2 ) or the wild-type EBV1 genome ( NC_007605 . 1 ) using Bowtie ( -v 3 –m 10 ) [17 , 18 , 82] . Mapped reads were analyzed by PARalyzer [83] to define 3’UTR interaction sites for EBV BART2 , 9 , 17 , 18 , and BHRF1-2 miRNAs . A comprehensive list of cellular genes associated with BCR signaling was curated from six publicly available databases ( Reactome , Panther Pathways , NDeX , KEGG , PathCard , and NetPath ) , and gene identifiers of the EBV miRNA 3’UTR interactions were compared to determine targets related to BCR signaling . Pathways interactions were compiled and drawn in PathVisio [84 , 85] . All luciferase and PCR data are reported as mean with standard deviations ( S . D . ) and values are derived from at least three independent experiments , unless otherwise stated , with technical replicates . Statistical significance was determined by paired Student’s t test , performed using Microsoft Excel 2010 , and values p < 0 . 05 were considered significant . | Understanding the molecular mechanisms regulating EBV latency and entry into the lytic replication cycle is important in developing therapies for viral disease . We demonstrate here that a subset of EBV miRNAs target host factors within the BCR signaling pathway and consequently , negatively regulate cellular responses to BCR cross-linking . Disrupting activity of individual EBV miRNAs , specifically miR-BHRF-1-2-5p and miR-BART2-5p , enhanced lytic gene expression following BCR engagement , suggesting a key role for these miRNAs in attenuating viral reactivation . Our experiments establish a link between EBV miRNAs and signaling through the BCR , providing new insight into miRNA-mediated aspects of latency and reactivation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"colorimetric",
"assays",
"medicine",
"and",
"health",
"sciences",
"luciferase",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"natural",
"antisense",
"transcripts",
"gene",
"regulation",
"enzymes",
"immunology",
"pathogens",
"microbiology",
"enzymology",
"biochemical",
"analysis",
"viruses",
"immune",
"receptor",
"signaling",
"micrornas",
"enzyme",
"assays",
"membrane",
"receptor",
"signaling",
"dna",
"viruses",
"mtt",
"assay",
"bioassays",
"and",
"physiological",
"analysis",
"herpesviruses",
"research",
"and",
"analysis",
"methods",
"white",
"blood",
"cells",
"epstein-barr",
"virus",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"gene",
"expression",
"oxidoreductases",
"microbial",
"pathogens",
"viral",
"replication",
"biochemistry",
"rna",
"antibody-producing",
"cells",
"signal",
"transduction",
"cell",
"biology",
"nucleic",
"acids",
"b",
"cells",
"viral",
"persistence",
"and",
"latency",
"virology",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"non-coding",
"rna",
"cell",
"signaling",
"organisms"
] | 2019 | Epstein-Barr virus microRNAs regulate B cell receptor signal transduction and lytic reactivation |
Replicating circular RNAs are independent plant pathogens known as viroids , or act to modulate the pathogenesis of plant and animal viruses as their satellite RNAs . The rate of discovery of these subviral pathogens was low over the past 40 years because the classical approaches are technical demanding and time-consuming . We previously described an approach for homology-independent discovery of replicating circular RNAs by analysing the total small RNA populations from samples of diseased tissues with a computational program known as progressive filtering of overlapping small RNAs ( PFOR ) . However , PFOR written in PERL language is extremely slow and is unable to discover those subviral pathogens that do not trigger in vivo accumulation of extensively overlapping small RNAs . Moreover , PFOR is yet to identify a new viroid capable of initiating independent infection . Here we report the development of PFOR2 that adopted parallel programming in the C++ language and was 3 to 8 times faster than PFOR . A new computational program was further developed and incorporated into PFOR2 to allow the identification of circular RNAs by deep sequencing of long RNAs instead of small RNAs . PFOR2 analysis of the small RNA libraries from grapevine and apple plants led to the discovery of Grapevine latent viroid ( GLVd ) and Apple hammerhead viroid-like RNA ( AHVd-like RNA ) , respectively . GLVd was proposed as a new species in the genus Apscaviroid , because it contained the typical structural elements found in this group of viroids and initiated independent infection in grapevine seedlings . AHVd-like RNA encoded a biologically active hammerhead ribozyme in both polarities , and was not specifically associated with any of the viruses found in apple plants . We propose that these computational algorithms have the potential to discover novel circular RNAs in plants , invertebrates and vertebrates regardless of whether they replicate and/or induce the in vivo accumulation of small RNAs .
Viroids and a group of satellite RNAs ( satRNAs ) have single-stranded circular RNA genomes that range in size from 220 to 457 nucleotides ( nt ) [1]–[4] . These subviral pathogenic RNAs lack protein-coding capabilities and thus depend on either host-encoded DNA-dependent RNA polymerase ( viroids ) or helper virus-encoded RNA-dependent RNA polymerase ( circular satRNAs ) for replication [5] , [6] . Viroids and circular satRNAs have been proven to be excellent biological models for studying non-coding RNAs ( ncRNAs ) and basic biology [3] , [4] . The most notable examples include the discovery of RNA-directed DNA methylation ( RdDM ) in viroid-infected plants [7] and of the hammerhead ribozymes in viroids [8] and circular satRNAs [9] of plant viruses . Interestingly , recent studies have revealed the production of thousands of non-replicating circular RNAs ( circRNAs ) across species from Archaea to humans [10] , [11] . These circRNAs are largely generated from back-spliced exons , in which splice junctions are formed by an acceptor splice site at the 5' end of an exon and a donor site at the downstream 3' end [10] , [12] , [13] . The functions of circRNAs are largely unknown , although a few circRNAs have recently been shown to play regulatory roles as , for example , microRNA sponges [10] , [12] , [13] . Viroids infect many crops and cause severe symptoms in susceptible hosts that result in economically important diseases [2] . However , the rate of discovery of the replicating circular RNAs is slow compared to the discovery of viruses [14] . For example , fewer than 40 viroid species , all of which infect higher plants , have been identified [15] since the first report in 1971 [1] . The slow rate of viroid discovery is often attributed to the technical difficulty involved in the purification and characterization of the naked non-coding circular RNAs that generally occur at low concentrations in the infected host [16] . We have recently described an approach for sequence homology-independent discovery of replicating circular RNAs by analyzing the total small RNA populations from samples of diseased tissues with a program known as progressive filtering of overlapping small RNAs ( PFOR ) [17] . The PFOR approach relies on the observations that rolling-circle replication of viroids and some satRNAs produces multimeric head-to-tail dsRNAs [5] and that continuous overlapping sets of small interfering RNAs ( siRNAs ) processed by Dicer [18]–[21] from the direct repeat dsRNAs accumulate to high levels in infected plant tissues [22] , [23] . PFOR retains viroid-specific siRNAs for genome assembly by progressively eliminating non-overlapping small RNAs and those that overlap but cannot be assembled into a direct repeat RNA . Use of PFOR for the analysis of a grapevine small RNA library led to the discovery of a viroid-like circular RNA of 375 nt that encodes active hammerhead ribozymes in both plus and minus polarities [17] . However , it remains unknown whether the identified circular RNA can initiate independent infection . Two major limitations of the first version of PFOR restrict its application in the discovery of circular RNAs . First , the iterative filtering of small RNAs that are not derived from a replicating circular RNA is a slow process and takes up more than 90% of the PFOR running time . Because PFOR was written in the explanatory PERL language , analyzing complex small RNA libraries using PFOR may take hours or days . Second , circular RNAs are not identifiable by PFOR if they neither replicate nor trigger Dicer-dependent siRNA production in a eukaryotic cell . In this study , a considerably improved version of PFOR was developed by adopting parallel programming in the C++ language . The use of the new version of PFOR , designated PFOR2 , led to the discovery of a new viroid from grapevine and a viroid-like RNA from apple tree . Moreover , a new program was developed and incorporated into PFOR2 for the discovery of distinct classes of circular RNAs , including those that neither replicate nor induce in vivo accumulation of Dicer-dependent siRNAs . We propose that the application of PFOR2 would speed up the discovery of novel circular RNAs and expand the list of known host species that can be independently infected by viroids .
The computational algorithm of PFOR has been developed to identify replicating circular RNAs including viroids by deep sequencing of small RNAs [17] . A key step of PFOR algorithm is to separate terminal small RNAs ( TSRs ) from internal small RNAs ( ISRs ) in a small RNA pool . A small RNA is defined as an ISR if it overlaps at least one other small RNA at both ends larger than k-mer in the pool , whereas a TSR overlaps at least one other small RNA larger than k-mer in the pool at only one end of the TSR . The process of PFOR includes two main steps: filtering all non-overlapping small RNAs and terminal small RNAs ( TSRs ) with overlapping k-mers and assembling the remaining internal small RNAs ( ISRs ) predicted to derive from circular RNAs ( Fig . 1C ) . Filtering TSRs derived from linear non-repeat precursor RNAs takes up more than 90% of PFOR running time . Therefore , to shorten the computing time required for filtering TSRs and to improve the performance of PFOR , PROR2 was developed by converting the previous algorithm written in the explanatory PERL language into the C++ language and adopting the parallel programming technology of OpenMP [24] . Because multiple shared memory filtering processes were performed concurrently in PFOR2 , the TSR filtering process was expected to be faster than PFOR ( Fig . 1A ) . To test the performance of PFOR2 , two publicly available small RNA libraries , in which both known viroids and viroid candidates have been verified by RT-PCR and Northern-blot hybridization , were used to compare the running time between PFOR and PFOR2 . Hop stunt viroid ( HpSVd ) , Grapevine yellow speckle viroid ( GYSVd ) and Grapevine hammerhead viroid-like RNA ( GHVd RNA ) were each identified , and their full-length genomic RNA sequences were obtained by both PFOR and PFOR2 from the grapevine tree sRNA library , which contains 4 , 701 , 135 reads of 18–28 nt in length ( GEO accession no . GSM458928 ) . However , PFOR2 required only 67 seconds and was 3 . 3 times faster than PFOR . The second sRNA library was from a peach tree infected with Peach latent mosaic viroid ( PLMVd ) and contained 7 , 862 , 905 reads of 18–28 nt in length ( GEO accession no . GSM465746 ) . Both PFOR and PFOR2 were again able to identify PLMVd and to recover the complete sequence . Instead of 2 hours by PFOR using a default k-mer of 17 , PFOR2 required only 22 minutes . These results demonstrated that PFOR2 was indeed faster than PFOR in viroid discovery . PFOR2 was next applied to determine whether an apple plant with typical symptoms of apple scar skin disease contained new circular RNAs . The sRNA library was constructed from this apple plant , and 15 , 321 , 500 clean reads of 18–30 nt in length with a predominant size of 21 nt were obtained after deep sequencing ( S1 Figure ) . Two putative circular RNAs were identified from the apple tree sRNA library by both PFOR and PFOR2 , although PFOR2 was six times faster . The first RNA species was 333 nt and shared 96% sequence identity with a variant of Apple scar skin viroid ( ASSVd ) ( accession no . KC110858 ) isolated previously from apple in China and was hence considered to be a new isolate of ASSVd . The second RNA species was 434 nt in length and showed no sequence similarity to any of the known entries in GenBank . Interestingly , the second predicted circular RNA also contained the conserved sequences found in hammerhead ribozymes as shown previously for GHVd RNA [17] . Thus , the second RNA identified from the apple sRNA library by PFOR and PFOR2 was tentatively designated as apple hammerhead viroid-like RNA ( AHVd-like RNA ) . To verify the predicted sequence and the circular nature of AHVd-like RNA , total RNAs from the diseased apple were isolated for divergent RT-PCR analysis . According to the sequence of AHVd-like RNA assembled by PFOR2 , two pairs of adjacent primers with opposing polarities ( AHVd-13F/12R and AHVd-88F/87R , sequences of primers are shown in S1 Table ) were designed for RT-PCR so that the full-length AHVd-like sequences would be amplified only when AHVd-like RNA existed in a circular form . We found that RT-PCR analysis of the apple RNA sample with either primer pair yielded a single DNA species with the expected size , demonstrating the circular nature of AHVd-like RNA from the apple tree ( Fig . 2A ) . Moreover , direct DNA sequencing of the RT-PCR products confirmed the sequence of AHVd-like RNA assembled by PFOR2 . We noted that the full-length AHVd-like RNA could be amplified when either of the primer pairs was used in RT reactions , indicating the existence of both plus and minus circular RNA molecules in the infected tissue . To further investigate the in vivo properties of AHVd-like RNAs , nucleic acid preparations were analyzed by denaturing PAGE and Northern-blot hybridization with a probe either corresponding or complementary to the assembled full-length AHVd-like RNA . The characteristic circular and linear forms were detected in the infected tissue but not in the healthy apple plant ( Fig . 2B ) . This result further validated the in vivo circularity of AHVd RNAs . Furthermore , the AHVd-like RNAs with opposing polarities appeared to accumulate at different levels ( Fig . 2B ) . Given that the strand accumulating at a higher level is arbitrarily assigned to be the plus polarity , the sequence of AHVd-like RNA obtained by PFOR2 was designated as the plus strand . Cloning and sequencing of full-length cDNA clones of viroids would supply relevant information on sequence variability in the natural viroid-like RNA populations . The sequenced cDNA clones of AHVd-like RNA were amplified by the two primer pairs , AHVd-13F/12R and AHVd-88F/87R . Therefore , the putative mutations located at the positions of one pair of primers were determined through amplification and sequencing with the second pair of primers . In total , 14 sequences of AHVd-like RNA were obtained . None of these AHVd-like RNA sequences were 100% identical to other 13 sequences . However , one sequence ( clone of 1–12 shown in S2 Figure ) was identical to the assembled sequence of AHVd-like RNA by PFOR2 . The alignment of these 14 sequences revealed the presence of 36 mutations in the population of AHVd-like RNA . Although a high-fidelity DNA polymerase was used for PCR amplification , we were not able to exclude possible errors introduced during RT-PCR . Thus , after 22 mutations detected only in one clone were removed , the remaining 14 mutations found in at least two clones were tentatively considered to be natural variations ( Fig . 3A and S2 Figure ) . The above analyses showed that the clone of 1–12 represented consensus sequences of AHVd-like RNA , a circular molecule of 434 nt consisting of 114 G ( 26 . 3% ) , 116 C ( 26 . 7% ) , 96 A ( 22 . 1% ) , and 108 U ( 24 . 9% ) with a G+C content of 53% ( Fig . 3A ) . AHVd-like RNA did not contain the characteristic central conserved region ( CCR ) found in the viroid family Pospiviroidae [15] . However , both strands of AHVd-like RNA could be folded into the conserved hammerhead ribozyme structure found in the Avsunviroidae and other small catalytic RNAs [25] . The predicted secondary structure of minimal free energy for AHVd-like RNA was of the quasi-rod-like class of viroids and showed two bifurcations at the right terminal part of the molecule ( Fig . 3A ) , which was similar to that of Eggplant latent viroid ( ELVd ) [26] . The paired nucleotide residues represented 68 . 2% of the total , including 56 . 8% G:C , 35 . 1% A:U , and 8 . 1% G:U . Interestingly , 11 out of 14 observed mutations either were mapped in the loop regions or did not affect base pairing ( Fig . 3A ) , which indirectly supported the proposed secondary structure of AHV-like RNA existing in vivo . Both strands of AHVd-like RNA could form natural hammerhead structures ( Fig . 3B ) containing 11 strictly conserved residues [27] and the adjacent helices flanking the self-cleavage sites of a group of small catalytic RNAs . In the plus and minus hammerhead structures of AHVd-like RNA , helix III was stable and helices I and II were closed by short loops 1 and 2 . These features were similar to the hammerhead structures of ( i ) ELVd [26] , ( ii ) PLMVd [27] , ( iii ) satellite RNAs of the nepoviruses and sobemoviruses [9] , [28] , ( iv ) a cherry small circular RNA ( csc RNA1 ) [29] , and ( v ) GHVd RNA discovered very recently in grapevine [17] . The hammerhead structures of AHVd-like RNA were carefully compared with those of other known viroids and circular satRNAs ( namely viroid-like RNAs ) , revealing some common salient features ( Fig . 3B ) . i ) In most natural hammerhead structures , positions 10 . 1 and 11 . 1 form a G-C pair , and positions 15 . 2 and 16 . 2 form a C-G pair ( see ref . [30] for nomenclature ) . AHVd-like RNA hammerhead structures conformed to this rule . ii ) A cytidylate residue preceded the predicted self-cleavage sites of AHVd RNA hammerhead structures , as occurs in most other known hammerhead structures . iii ) The residue of position 7 between the conserved CUGA and GA sequences was a U in both RNA hammerhead structures of AHVd-like , which also conformed to the examples observed in most natural hammerhead structures wherein this residue is U , C , and , exceptionally , A . However , the hammerhead structures of AHVd-like RNA exhibited some peculiarities . Both hammerhead structures of AHVd-like RNA shared sequence similarities in the helices and loops with the strictly conserved helix II and loop 1 ( Fig . 3B ) . Sequence similarities included 4 base-pairs of CAGG with CCUG , forming helix II in the consensus hammerhead structure of AHVd-like RNA ( Fig . 3B ) , which corresponded to the equivalent positions in the consensus hammerhead structure of ELVd , the plus strand hammerhead structures of GHVd RNA [17] , satellite RNAs of Chicory yellow mottle virus ( CYMV ) and Tobacco ringspot virus ( TRSV ) [31] . Moreover , loop 2 of the ( + ) hammerhead of AHVd-like RNA contained 7 nucleotides and was the largest reported among natural hammerheads . Importantly , the base substitutions found in different AHVd-like RNA variants in the region of the hammerhead structures did not affect the stability of–helix III and no mutations were found in helices I and II ( Fig . 3B ) , suggesting that these self-cleaving domains were functional in vivo . The activity of the predicted ribozymes encoded by AHVd-like RNA was investigated . Full-length monomeric plus and minus AHVd-like RNA transcripts were synthesized in vitro from linearized plasmids and were found to be self-cleaved during transcription and after purification when incubated under standard self-cleavage conditions in a protein-free buffer ( Fig . 4 ) . The cleaved fragments ( 5′F and 3′F ) for the plus and minus strands of the transcripts showed the expected lengths based on the predicted self-cleavage sites of the hammerhead structures ( Fig . 4 ) . The predicted cleavage sites ( Fig . 3A and 3B ) were also experimentally confirmed by rapid amplification of 5′-cDNA ends ( 5′-RACE ) -PCR ( S3 Figure ) . We noted that the plus strand full-length AHVd-like RNA transcripts were more stable during transcription than the minus strand transcripts ( Fig . 4B ) , suggesting a higher self-cleavage efficiency of the minus strand hammerhead ribozyme . Although the above analyses determined the circularity and self-cleavage activity of AHVd-like RNA , it was still unclear whether AHVd-like RNA represented a new viroid or a circular satRNA . If AHVd-like RNA corresponded to a new plant circular satRNA , it was expected that a helper virus would be present in the diseased tissues to support its replication . To this end , the sRNAs from the diseased apple tree were assembled by Velvet program [32] . BLAST analysis identified contigs that showed sequence similarities with Apple chlorotic leaf spot virus ( ACLSV ) , Apple stem grooving virus ( ASGV ) , and Apple stem pitting virus ( ASPV ) . The presence of these three plant viruses was further confirmed by RT-PCR ( S4 Figure ) . However , we noted that none of these three plant viruses have been reported to have satRNAs . A survey of 182 apple tree samples from different cultivars was performed to determine whether AHVd-like RNA co-existed with any of these viruses . We found that AHVd-like RNA was detected in 75 of the 182 apple tree samples . Notably , the incidence of AHVd-like RNA was not associated with ACLSV , ASPV , or ASGV , suggesting that AHVd-like RNA might be a novel viroid ( S2 Table ) . However , the viroid nature of AHVd-like RNA remained to be verified because neither Northern-blot hybridization nor RT-PCR detected the replication of AHVd-like RNA in the apple seedlings free of AHVd-like RNA one year after mechanical inoculation with the dimeric transcripts synthesized in vitro from the full-length cDNA clones of AHVd-like RNA described above . Given that the size distribution of sRNAs derived from viroids might serve as an indicator of the subcellular localization or replication sites of the viroids [33] , [34] , we next compared the accumulation and profile of sRNAs derived from AHVd-like RNA with those of ASSVd in the same tissues of the apple tree . Similar to ASSVd , AHVd-like RNA specific sRNAs from different size families were all divided approximately equally into the plus and minus strands and the predominant sRNA species was the 21 nt class ( Fig . 5A ) . However , few of the AHVd-like RNA-specific sRNAs were 24 nt long . A similar size distribution profile was observed for vd-sRNAs from tissues infected by PLMVd [33] , [35] and GHVd [17] . In contrast , a notable amount of ASSVd sRNAs belonged to the 24 nt class ( Fig . 5B ) , similar to several viroids that replicate in the nucleus [34] , [36]–[38] . These findings suggest that AHVd-like RNA may not replicate in the nucleus . AHVd-like RNA specific sRNA reads of 21 to 24 nt in length were mapped to the corresponding positions on the AHVd-like RNA genomic or anti-genomic RNAs ( Fig . 5C ) . As previously reported for PLMVd sRNAs isolated from infected peach [30] , [32] and GHVd RNA-specific sRNAs from infected grapevine [17] , the sRNAs of AHVd-like RNA were derived from every position in both the genomic and anti-genomic strands , and their distribution was biased , with a profile of several hotspots ( Fig . 5C ) . Grapevine is a natural host for many viroids [39] . Although most of these viroids do not induce symptoms in grapevine , cultivated grapevines with latent viroid infections may serve as reservoirs for certain viroids to infect crops and cause severe diseases [40] . The discovery of a novel viroid-like circular RNA from the original ‘Pinot noir’ grapevine by PFOR [17] suggests that more novel viroids or viroid-like RNAs may exist in cultivated grapevines , especially in some old grapevines . Collections of several grapevine stocks of at least 100 years of age in Xinjiang , China [39] allowed us to test this hypothesis . Of these grapevine trees , a ‘Thompson Seedless’ plant grown in Tulufan was selected for sRNA deep sequencing and viroid discovery by PFOR2 . The obtained sRNA library contained 14 , 033 , 487 clean reads of 17–30 nt in length , with 21 nt as the most dominant size class ( S5 Figure ) . PFOR2 analysis of the library took 2 hours and 24 minutes and was 7 . 1 times faster than PFOR analysis . Complete genomes of four known viroids: HpSVd , GYSVd-1 , GYSVd-2 , and Australia grapevine viroid ( AGVd ) , which have been previously detected by RT-PCR and Northern-blot hybridization in this old grapevine tree [39] , were assembled by PFOR2 . PFOR2 analysis of the grapevine sRNA library also identified a putative circular RNA molecule of 328 nt in length that shared 79% sequence similarity with Citrus viroid VI ( CVd-VI ) ( accession no . AB019508 ) . Because CVd-VI had not been previously isolated from grapevine and the sequence similarity between CVd-VI and the identified RNA molecule was below the viroid species demarcation criteria of 90% sequence similarities [15] , we hypothesized that the circular RNA revealed by PFOR2 represented a new viroid and was tentatively designated as Grapevine latent viroid ( GLVd ) hereafter . To confirm the viroid nature of GLVd , we first determined whether GLVd existed in a circular form in vivo . Two sets of adjacent primers of opposite polarity ( GLVd-252F/251R and GLVd-141F/140R , shown in S1 Table ) , derived from the predicted sequence by PFOR2 , were synthesized and used for the amplification of the full-length circular GLVd by RT-PCR . As a control , PCR was performed with these primers using the template of total DNA isolated from the old grapevine without the RT step to determine whether GLVd was derived from repeat elements of the host genome . Divergent RT-PCR with either of the two primer pairs yielded a product with the expected size whereas no specific products were amplified by PCR ( Fig . 6D ) , confirming the circular RNA nature of GLVd . The amplified DNA of the expected size was eluted , and four clones from each primer pair were sequenced . Sequence analysis revealed the presence of a master sequence represented by six clones , while the two sequence variants contained a deletion of A at position 54 and a substitution ( G/A ) at position 125 , respectively ( Fig . 6A ) . Importantly , the master sequence of GLVd was identical to the sequence discovered by PFOR2 and was 328 nt in length , with 67 A ( 20 . 4% ) , 70 U ( 21 . 3% ) , 96 G ( 29 . 3% ) and 95 C ( 29% ) , producing a G+C content of 58 . 3% . The availability of the full-length GLVd genomic sequence allowed us to synthesize a GLVd-specific riboprobe for detecting various molecular forms of GLVd RNA by Northern-blot hybridization . Total RNAs extracted from the old grapevine were separated by denaturing PAGE followed by Northern-blot hybridization , leading to the detection of the characteristic circular and linear forms ( Fig . 6D ) . These findings together indicated that GLVd existed as a circular RNA in the old grapevine tree . The minimal free-energy secondary structural prediction revealed a rod-like conformation of GLVd . The predicted secondary structure of GLVd was similar to that proposed for most viroids [15] , [41] and contained 63 . 4% paired nucleotides , including 67 . 3% , 24 . 0% and 8 . 7% of G:C , A:U , and G:U pairs , respectively ( Fig . 6A ) . Notably , the GLVd structure contained the central conserved region ( CCR ) , which is the key structural element and taxonomic criterion for assigning viroids to the family Pospiviroidae . The sequences of upper and lower CCR of GLVd were nearly identical to that of Apple scar skin viroid ( ASSVd ) [42] , the type species of the genus Apscaviroid . The terminal conserved region ( TCR ) of GLVd was also similar to that found in apscaviroids ( Fig . 6A and B ) . Furthermore , the GLVd structure included a polypurine stretch located in the pathogenicity domain , which is conserved in the family Pospiviroidae [41] . Hairpin I ( HPI ) formed by the upper CCR strand and the flanking inverted repeat [43] , [44] is a conserved structural element in the family Pospiviroidae . A typical HPI was identified in GLVd and included the capping palindromic tetraloop , the adjacent 3-bp stem , and the 7-bp stem interrupted by two opposite-facing nucleotides that were seemingly unpaired ( Fig . 6C ) . However , sequence variations were noted in the HPI of GLVd compared to the known apscaviroids ( Fig . 6C and S6 Figure ) . The nucleotide substitution of U by C at the left terminus of the upper CCR converted a G:U base-pair in the stem of HPI into a G:C base-pair , which was predicted to strengthen the stability of this structure . In contrast , a large internal loop present in GLVd HPI would weaken the stability of HPI ( Fig . 6C ) . Detection of the conserved structural features of viroids such as CCR , TCR , and HPI in GLVd further supports the idea that GLVd is a viroid . To further verify the viroid nature of GLVd , dimeric head-to-tail transcripts of GLVd were transcribed in vitro from the constructed cDNA clones of GLVd . Virus-free grapevine seedlings ( cv ‘Beta’ ) grown in early spring were mechanically inoculated with the GLVd transcripts by slashing the stems with razor blades . Uninoculated healthy seedlings from the same batch were kept as controls . Because GLVd was undetectable by Northern-blot hybridization 3 and 6 months post inoculation , we re-inoculated the seedlings with a higher dose of GLVd transcripts and detected weak hybridization signals from 4 of the 18 inoculated grapevine plants 3 months after the secondary inoculation . To facilitate GLVd detection in the young tissues , the apical shoots of the inoculated grapevine plants were removed , and the leaves from the young lateral branches that emerged 6 months after the secondary inoculation ( or 12 months after the first inoculation ) were collected for RNA extraction . We found that GLVd infection became readily detectable in 6 of the 18 inoculated grapevine plants using either Northern-blot hybridization or RT-PCR . The progeny sequence was determined via DNA sequencing of the cloned RT-PCR products and was found to be the same as the inoculated GLVd transcripts . Therefore , GLVd autonomously replicated in its natural host grapevine , fulfilling the most critical criterion to be considered as a viroid . To determine the taxonomy of GLVd , the sequence of GLVd was aligned with all of the known species in the genus Apscaviroid . The phylogenetic tree constructed using viroids of genus Colviroid as the out-group revealed two subgroups of apscaviroids ( Fig . 7A ) . GLVd was clustered in subgroup-II and most closely related to CVd-VI and a tentative new species of Persimmon viroid 2 ( PVd-2 ) identified very recently from American persimmon ( Diospyros virginiana L . ) [45] ( Fig . 7A ) . These results indicated that GLVd should be considered as a new member in the genus Apscaviroid . Interestingly , careful inspection of the apscaviroid alignments identified two types of repeat sequences between GLVd and PVd-2 ( Fig . 7B ) . Further study is necessary to determine whether the repeat sequences were involved in host adaptation because simple sequence repeats ( SSRs ) distribute non-randomly in viroid genomes and might play a significant role in the evolution of viroids [46] . We next developed a simple computational program , Splitting Longer reads into Shorter fragments ( SLS ) , as part of PFOR2 to discover biologically active circular RNAs via the deep sequencing of long RNAs instead of small RNAs . The program cut the sequenced long RNAs into 21-nt virtual sRNAs of 20-nt overlap with their 5′ and 3′ neighboring sRNAs before PFOR2 analysis to retain only 21-nt virtual ISRs for the final assembly of circular RNAs ( Fig . 1B ) . To determine the efficacy of SLS-PFOR2 , we sequenced the total RNAs from PSTVd-infected potato seedlings by constructing independent libraries using Not Not So Random ( NNSR ) library protocol [47] after either depletion of the abundant ribosomal RNAs [48] , [49] or enrichment for circular RNAs following the degradation of linear RNAs by RNase R [50] ( S7 Figure ) . The sequencing of the rRNA-depleted library yielded 774 , 621 reads longer than 100 nt , among which 83 reads were derived from PSTVd with a mean length of 160 nt . A total of 92 , 093 reads longer than 100 nt were obtained from the RNase R-treated library , with 55 reads from PSTVd . We found that the full-length PSTVd molecule of 354 nt was readily identified by SLS-PFOR2 from both the rRNA-depleted library ( k-mer 19 or 20 ) and the RNase R library ( k-mer 17 or 18 ) with a running time of 3 hours 20 minutes and 103 hours 14 minutes , respectively . These results demonstrated that SLS-PFOR2 is capable of discovering circular RNAs independently of the in vivo production or the deep sequencing of their specific small RNAs .
Next-generation sequencing ( NGS ) approaches can readily identify viral and subviral pathogens in samples of plant and animal diseased tissues that are related in nucleotide sequence or encoded protein sequence to a known pathogen . The development of PFOR for viroid discovery thus represents a conceptual advance because , unlike NGS and several available classical approaches , PFOR does not depend on sequence homology with a known viroid . The major improvements described in this study overcame the limitations of the published version of PFOR that restrict its potentially widespread applications in pathogen discovery . PFOR2 was 3 . 3 , 5 . 4 , and 7 . 1 times faster than PFOR in the analysis of the three small RNA datasets from grapevine , peach , and apple , respectively . The enhanced speed is likely to be critical for viroid discovery when targeting hosts with large genome sizes and/or abundant small RNA populations . For example , our analysis of a small RNA library from Areca catechu with 46 , 637 , 488 reads took 8 hours and 40 minutes by PFOR2 instead of 110 hours by PFOR ( unpublished data ) . The efficacy of PFOR2 was verified with the discovery of GLVd as a novel viroid that initiates independent infection in its natural host . Moreover , the development of SLS-PFOR2 eliminates the requirement for the in vivo production and accumulation of Dicer-dependent siRNAs to target the circular RNAs to be identified . As a result , small RNA sequencing becomes unnecessary , and RNA-seq libraries depleted of either ribosomal RNAs or linear RNAs can be analyzed by SLS-PFOR2 for the discovery of both replicating and non-replicating circular RNAs in diverse organisms . In principle , SLS-PFOR2 can identify novel viroid circular RNAs in host species that replicate but do not trigger the biogenesis of viroid-specific siRNAs . Therefore , SLS-PFOR2 has the potential to expand the list of both viroids and host species that support viroid infection . PFOR and FOR2 separate small RNAs in the pool into TSR and ISR groups based on the presence of the minimal overlapping k-mer among reads and remove all TSRs progressively . As a result , the overlapping sets of small RNAs retained after the filtering process might be different when different k-mers are used , leading to the variations in the sequences assembled by PFOR and PFOR2 that may not reflect the natural heterogeneity of viroids . The successful detection of each viroid by PFOR and PFOR2 is dependent on whether the circular RNA has been completely covered by a set of overlapping small RNAs with the minimal length defined by k-mers at both ends after removing all TSRs . Because each ISR is allowed to be used only once during the assembling step , only one viroid would be revealed when two or more viroids share small RNAs with lengths defined by k-mers or longer . For example , although ASSVd was revealed by PFOR2 analysis of the apple library using k-mers in the size range of 18 to 20 nt , AHVd-like RNA was identified using k-mers of 18 or 19 , but not of 20 . In the grapevine sample , HpSVd was identified by PFOR2 using k-mers of 17 to 21 , whereas GYSVd-1 and GYSVd-2 were each revealed with a specific k-mer most likely because the two viroids are 80% identical in sequence and share small RNAs . Therefore , it is necessary to analyze each small RNA library using PFOR2 with k-mers from 17 to 21 and to verify the assembled sequences of viroid candidates by RT-PCR and cDNA sequencing . The evolutionary origin of viroids remains unknown [51] . However , it has been proposed that most , if not all , present viroid diseases of cultivated plants originated recently by the accidental introduction of viroids from endemically infected wild plants into susceptible cultivated plants [52] . Thus , the identification of the original wild host plants as symptomless viroid carriers may provide additional insight into possible evolutionary scenarios . Cultivated grapevines were assumed to be associated with ‘Etrog’ citron fruit , displaying citrus viroid disease symptoms as depicted in an ancient synagogue from the early 6th century C . E . in Israel [40] . This suggested that cultivated grapevines with latent infections of viroids may serve as reservoirs for viroid spreading and causing diseases in other hosts . Accordingly , the viroids that cause some epidemic diseases at present are likely to come from the originally infected grapevines . This hypothesis is consistent with the finding that the cultivated grapevines asymptomatically infected with HpSVd were considered as the origin of the hop stunt disease epidemic in commercial hops in Japan [53] . It is also possible that grapevines might harbor some unknown viroids that are yet to be identified . The discovery of a novel viroid-like circular GHVd RNA previously [17] and GLVd in this work supports this idea . GLVd is related to both CVd-VI and PVd-2 , which were isolated from diseased Etrog citrons ( Citrus medica L . ) with mild petiole necrosis and leaf bending [54] and American persimmon [45] , respectively . Since GLVd and PVd-2 appear to originate from a common ancestor , it will be of interest to determine in future studies if the two repeated sequences detected between GLVd and PVd-2 ( Fig . 7B ) play a role in host adaptation during transmission from its original host to certain new susceptible hosts . Our conclusion that GLVd is a novel viroid is supported by the molecular and biological evidence presented here including its circularity , typical structural elements of viroids , and self-replication in grapevine seedlings . The phylogenetic analysis indicates that GLVd is a member of the genus Apscaviroid . Although we found that GLVd was able to independently infect grapevine seedlings ( cv ‘Beta’ ) , the accumulation of GLVd was low , and no obvious symptoms were observed in infected grapevine plants . Although future studies on biological properties of GLVd may further differentiate this viroid from those previously reported , the conserved structural elements , the low sequence identity ( maximum of 79% with CVd-IV ) with other members in the genus Apscaviroid , and the natural host of GLVd , strongly support the possibility of annotating it as a new species in the genus Apscaviroid . It is currently unclear whether AHVd-like RNA is a viroid or a satellite RNA , in contrast to GLVd . AHVd-like RNA shared no homology with the apple genome and was not amplified by PCR without a RT step , indicating that AHVd-like RNA was exogenous . Given that AHVd-like RNA encoded self-cleavage activities and was not specifically associated with any of the viruses identified in apple trees , we propose that AHVd-like RNA is a viroid . However , we were unable to demonstrate independent infectivity in apple seedlings for either the in vitro transcripts from dimeric AHVd-like RNA cDNA clones or AHVd-like RNA purified from apple tissues . In this regard , AHVd-like RNA may be related to ASSARNA-2 , a circular RNA that was previously isolated from diseased apple plants in Japan and China , known to migrate more slowly than the 330-nt ASSVd RNA and unable to establish independent infection in apple seedlings [55]–[57] . Furthermore , our search for the conserved tertiary structure of a kissing loop , which is found in most Avsunviroidae viroids [58] , [59] and in GHVd RNA [17] , identified only weak kissing loops of 3 base-pairs in AHVd-like RNA ( S8 Figure ) . Therefore , we cannot rule out the possibility that AHVd-like RNA is a novel satRNA . However , we note that virus-derived siRNAs produced by the antiviral Dicer of a fungal host are predominantly within the 20- to 22-nt range with a peak at 21-nt [60] . It is therefore less likely that AHV-like RNA replicates and triggers Dicer recognition in a fungal host since 21- and 22-nt small RNAs derived from AHVd-like RNA were clearly more abundant than 20-nt and the remaining size classes of small RNAs as found for plant viral siRNAs produced hierarchically by Dicer-like 4 ( DCL4 ) and DCL2 [61] .
For the initial identification of viroids and viroid-like RNAs from apple , leaves were collected from an apple ( Malus pumila Mill . cv . Fuji ) plant , the fruits of which showed typical symptoms of apple scar skin viroid disease , in Shandong province China , in July 2012 . The grapevine ( Vitis vinifera L . ) leaf samples used for determination of GLVd were from Tulufan in Xinjiang China . This grapevine plant ( cv . Thompson seedless ) for sample collections is more than 100 years old . Young leaves of both apple and grapevine were immediately put into RNAlater stabilization solution ( Ambion , USA ) after collection and sent to a laboratory for deep sequencing analysis . Moreover , approximately 10 g of apple and grapevine leaves were packaged with ice , kept fresh at low temperature , and sent to a laboratory for RNA analysis using PAGE and northern-blot hybridization . To survey the occurrence of viroid-like apple RNA in China , 182 leaf samples of variant apple cultivars from five provinces were collected from 2012 to 2014 and kept at −80°C . Total RNAs used for deep sequencing analysis were extracted by TRIzol reagent ( Invitrogen ) following the manufacturer's instructions . The integrity of the resulting RNA preparations was evaluated before preparing cDNA libraries using an Agilent Technologies 2100 bioanalyzer . For RNA analysis by PAGE and northern-blot hybridization , nucleic acid preparations were obtained with buffer-saturated phenol followed by ethanol precipitation , as reported previously [62] . Methoxyethanol and CTAB were used to remove polysaccharides during purification [62] . To prepare templates of RT-PCR amplification performed for cloning full-length sequence of viroid-like apple RNA , the obtained crude extracts were run on a non-denaturing 5% polyacrylamide gel stained with ethidium bromide , and the region of the gel delimited by the 250-bp and 400-bp DNA markers was excised and eluted as described previously [26] . In the experiment involving the RNA-seq of the potato samples , the extracted total RNAs by TRIzol reagent were purified by depleting ribosomal RNAs and non-circular RNAs . RNA species smaller than 200 nt , such as 5S ribosomal RNA , were first removed using the RNeasy Mini Kit ( Qiagen , USA ) , and 28S and 18S ribosomal RNAs were depleted by hybridization with specific probes following the instructions for the RiboMinus Plant Kit ( Invitrogen , USA ) . To enrich circular RNAs , the total RNAs from the same sample were digested with RNase R ( Epicentre , USA ) at 37°C for 90 min to remove non-circular RNAs . The RNA extracts were separated using two-dimensional PAGE ( 2D-PAGE ) under non-denaturing and denaturing conditions and stained with ethidium bromide , as previously described [63] . To determine the circularity of RNAs , the total RNAs were run on denatured PAGE gel containing 8 M urea and then transferred to Hybond N+ nylon membranes by upward capillary transfer in 20×SSC buffer . The hybridization was performed at 68°C overnight by specific probes that were generated by a DIG RNA labeling kit ( Roche ) according to the manufacturer's instructions . The immunological detection was performed by adding chemiluminescent substrate to the membrane following the manufacturer's instructions . The small RNA libraries were constructed using Illumina's small RNA sample preparation Kit ( Invitrogen , USA ) following Illumina's method . The protocols for sRNA purification , adaptor ligation , RT-PCR amplification , library purification and high-throughput DNA sequencing on an Illumina HiSeq-2000 have been reported previously [64] , [65] . Two sRNA libraries of an old grapevine plant and an apple tree were sequenced . Raw data from the Illumina platform were first processed to trim adaptor and barcode sequences . Reads of 18-30 nt were extracted from the obtained trimmed reads to generate sRNA libraries for assembly . The sRNA library of a peach tree infected with PLMVd ( accession no . GSM465746 ) and the small RNA library from a grapevine tree cultivar Pinot noir ENTAV115 ( accession no . GSE18405 ) were downloaded from the NCBI Gene Expression Omnibus ( GEO ) database . All of the prepared sRNA libraries were fed into an in-house pipeline . Briefly , exogenous sRNA was enriched by subtracting sRNA derived from the host genome using the Bowtie2 with default parameters [66] . The highly enriched exogenous siRNA from each sample were assembled de novo using Velvet [32] and PFOR [17]/PFOR2 . The resulting contigs were queried against the GenBank nt and nr databases using the BLAST program [67] . The RNA-seq libraries of potato samples were constructed with a modified Not Not So Random ( NNSR ) sequencing method [47] . Two libraries of potato were sequenced using an Ion-torrent sequencer according to the manufacturer's instructions . PFOR in the PERL language was converted to PFOR2 initially by using the C++ language . OpenMP is an Application Programming Interface ( API ) that supports multi-platform shared memory multiprocessing programming [24] . The parallel programming technology OpenMP was employed by PFOR2 to parallelize the filtering process of singletons and TSRs concurrently . Vector was also used in PFOR2 to store all sequences temporarily in the filtering process to simplify OPENMP parallelization . In PFOR , a two-level hash table was built at each iteration process to store all sequences in the pool , whereas in PFOR2 , a two-level hash table was only established at the first iteration , and non-ISRs were deleted from the two-level hash table for each subsequent iteration . The SLS ( Splitting Longer read into Shorter fragments ) program was developed to cut longer reads into virtual sRNAs . The final number of generated virtual sRNAs was dependent on two metrics: sRNA size and overlap size between neighboring sRNAs ( step size ) . Typically , a longer read was cut into contiguous 21-nt sRNAs covering the whole read , in which each sRNA overlapped 20 nt with its 5′ and 3′ neighboring sRNAs ( step size = 1 ) . The primers of HpSVd , GYSVd-1 , GYSVd-2 and AGVd were described previously [39] . The primer sets for amplification of GLVd and viroid-like apple RNA were designed from the corresponding sequences of contigs assembled by PFOR2 and were listed in Supplemental Table 1 . The first-strands of cDNAs were synthesized with Mu-MLV reverse transcriptase ( Promega ) at 42°C for 1 h , and PCR amplification was performed by high-fidelity pfu DNA polymerase ( Thermo , USA ) to generate full-length sequences of viroids and viroid-like RNA . The products of RT-PCR amplification were ligated with additional adenine ( A ) at the end using Taq DNA polymerase ( Takara , Dalian ) and cloned into pGEM-T vectors ( Promega ) with protruding 3′-terminal thymine ( T ) . The recombinant plasmids were amplified by transforming DH5α Escherichia coli cells , and positive clones were randomly selected for sequencing . The sequenced recombinant plasmids containing full-length of AHVd-like cDNA amplified with primers of AHVd-88F and AHVd-87R were digested with Nco I or Sal I to generate linear plasmids . RNA transcripts in both orientations were synthesized by T7 and SP6 RNA polymerase as described previously [29] , [68] . The products of in-vitro transcription were purified by RNeasy Mini Kit ( Qiagen , USA ) . The purified transcripts were incubated at 37°C for 1 h and then separated by 5% denaturing PAGE containing 8 M urea and visualized by ethidium bromide staining . Full-length of AHVd transcripts and the longer fragments resulting from their in vitro self-cleavage were excised from the gels and eluted , separately . The ribozyme activities of the purified transcripts were assessed according to previously described methods [29] , [68] . The purified longer fragments were used to validate the self-cleavage sites of AHVd-like RNA by 5′RACE amplification , which was conducted using the 5′RACE System for Rapid Amplification of cDNA Ends kit ( Invitrogen ) . Head-to-tail dimmers of the entire sequence of GLVd and AHVd-like RNA were prepared by ligation of unit-length inserts and cloning into pGEM-T vectors ( Promega ) , as described previously [69] . The orientation of the inserts of dimeric cDNAs was validated by sequencing . The resulting recombinant plasmids were digested into linear forms and used to synthesize dimeric transcripts with positive polarity by T7 RNA polymerase ( Promega ) . The dimeric transcripts of GLVd and AHVd-like RNA were mechanically inoculated into grapevine ( cv ‘Beta’ ) and apple virus-free seedlings ( cv ‘Fuji’ ) , respectively , by slashing the stems with razor blades . Each seedling was inoculated with at least 500 ng of dimeric transcripts . The inoculated seedlings were grown in a common greenhouse . The infectivity of the infectious clones of GLVd and AHVd-like RNA was examined by northern-blot hybridization every three months . The secondary structures with minimum free energy for GLVd and AHVd-like RNA were predicted by the circular version of the MFold program [70] . The obtained secondary structures were further edited for print by RnaViz 2 [71] . To search for possible kissing-loops in AHVd-like RNA , the Kinefold web server [72] was used with default parameters . | Viroids are a unique class of subviral pathogens found in plants , and they are difficult to identify since they are free circular non-coding RNAs and often replicate to low levels in host cells . We previously described the computational algorithm PFOR that discovers viroids by analyzing total small RNAs of the infected plants obtained by next-generation sequencing platforms . However , the algorithm written in PERL language is very slow , and viroid identification depends on the in vivo accumulation of extensively overlapping sets of small RNAs to target viroids . Here we report the development of PFOR2 that adopted parallel programming in the C++ language and was significantly faster than PFOR . We also describe a simple computational program that after incorporation into PFOR2 is capable of identifying viroids from deep sequencing of long RNAs instead of small RNAs . Moreover , we report the identification of Grapevine latent viroid ( GLVd ) and Apple hammerhead viroid-like RNA by the computational approach . Since our new algorithms do not depend on the analysis of viroid-derived small RNAs produced in vivo , it is possible to discover viroids in a wide range of host species including plants , invertebrates and vertebrates . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"science",
"viroids",
"plant",
"pathogens",
"plant",
"pathology",
"biology",
"and",
"life",
"sciences"
] | 2014 | Discovery of Replicating Circular RNAs by RNA-Seq and Computational Algorithms |
The transmission of hemorrhagic fever with renal syndrome ( HFRS ) is influenced by environmental determinants . This study aimed to explore the association between atmospheric moisture variability and the transmission of hemorrhagic fever with renal syndrome ( HFRS ) for the period of 1991–2010 in Changsha , China . Wavelet analyses were performed by using monthly reported time series data of HFRS cases to detect and quantify the periodicity of HFRS . A generalized linear model with a Poisson distribution and a log link model were used to quantify the relationship between climate and HFRS cases , highlighting the importance of moisture conditions . There was a continuous annual oscillation mode and multi-annual cycle around 3–4 years from 1994 to 1999 . There was a significant association of HFRS incidence with moisture conditions and the Multivariate El Niño–Southern Oscillation Index ( MEI ) . Particularly , atmospheric moisture has a significant effect on the propagation of HFRS; annual incidence of HFRS was positively correlated with annual precipitation and annual mean absolute humidity . The final model had good accuracy in forecasting the occurrence of HFRS and moisture condition can be used in disease surveillance and risk management to provide early warning of potential epidemics of this disease .
Hemorrhagic fever with renal syndrome ( HFRS ) is a rodent-borne disease caused by hantavirus , and is characterized by fever , hemorrhage , kidney damage , and hypotension . HFRS has been recognized as an important public health problem in Hunan Province , which is one of the most seriously affected areas in mainland China [1] , [2] . HFRS was first detected in Hunan Province in 1963 , and the highest incidence of the disease was recorded at 13 . 33/100 , 000 in 1985 . Changsha , the capital city of Hunan Province , also has a high prevalence of HFRS . Among Changsha areas , the highest incidence of HFRS was recorded in Ningxiang County , at 101 . 68/100 , 000 in 1994 . Many studies showed that the incidence rates of HFRS and the population dynamics of hosts are influenced by climatic factors , especially humidity . In Belgium , high soil moisture was significantly associated with the number of nephropathia epidemica ( NE ) cases [3] . In northeast China , relative humidity was positively associated with transmission of HFRS [4] . The highest HFRS incidence was recorded at 6 . 4/100 , 000 , in semi-humid areas with precipitation greater than 400 mm . In China , nearly 50% of HFRS cases have been in areas with precipitation greater than 800 mm [2] . Furthermore , under laboratory conditions , high humidity been proven to be beneficial for virus survival in the ex vivo environment [5] . Some studies also found that hantaviruses are limited in their spread to high-humidity environments for extended ex vivo stability [6] . However , little is known how the moisture condition , including seasonal variation and annual situation , influence the HFRS transmission in a relatively long period of time . In this study , we used wavelet analysis to investigate variations in dominant periodic cycles across the time series of HFRS , for quantifying the relationship between moisture condition and HFRS in the long run . We also aimed to develop a moisture condition-based forecasting model for the control and prevention of HFRS .
Changsha is a city with a total land area of 118 , 000 square kilometers and a population of about seven million in 2010 . Changsha comprises the Xiangjiang River alluvial plain , part of the Central Plain of China . The Xiangjiang , Weishui , Laodao , and Liuyang rivers and their branches flow through the city . These have promoted development of the region's traditional agricultural economy , with subtropical double-harvest rice cultivation of major importance . These environmental conditions provide the opportunity for HFRS to persist and spread ( Figure 1 ) . The present study was reviewed by the research institutional review board of Hunan CDC , and it is found that utilization of disease surveillance data did not require oversight by an ethics committee . The data were analyzed anonymously , using publicly available secondary data , therefore no ethics statement is required for this work . Records of HFRS cases from 1991 to 2010 were obtained from the the Hunan Notifiable Disease Surveillance System ( HNDSS ) . The data were from a passive surveillance system and all HFRS cases were first diagnosed by clinical symptoms , as defined by a national standard [7] . Clinical diagnosis criteria include: A person who had exposure to rodents and their feces , saliva and urine urine or who had traveled to the HFRS endemic area within two months prior to the onset of illness , and who had an acute illness with at least two of the following clinical symptoms: fever , chill , hemorrhage , headache , back pain , abdominal pain , acute renal dysfunction , and hypotension . Then , antibody-based serological tests were used ( e . g . MacELISA , IFA ) . A confirmed case of HFRS had to have had at least one of the laboratory criteria for diagnosis . Cumulative cases for each month over the study period were calculated to reflect seasonal fluctuations . In addition , the surveillance strategies for HFRS in Changsha have not changed during the study period . Monthly climatic data from 1991 to 2010 in Changsha were collected from the China Meteorological Data Sharing Service System ( http://cdc . cma . gov . cn/index . jsp ) . Climatic factors included monthly mean relative humidity and monthly accumulated precipitation . Absolute humidity was generated from a transformation of air pressure , relative humidity and temperature: ( 1 ) ( 2 ) ( 3 ) where ρw is absolute humidity , e is vapor pressure , Rv is the gas constant of water evaporation , L is the latent heat of water evaporation , es ( T ) is the saturation vapor pressure at the temperature T , φ is relative humidity and E is saturation vapor pressure . The multivariate El Niño–Southern Oscillation index was available from the Earth System Research Laboratory of the National Oceanic and Atmospheric Administration [8] . The index was used as an indicator of the global climate pattern on the six main observed variables over the tropical Pacific . These six variables are: sea-level pressure , zonal and meridional components of the surface wind , sea surface temperature , surface air temperature , and total cloudiness fraction of the sky . Wavelet analysis is suitable for investigating time series data from non-stationary systems [9] , [10] . Wavelet analyses have been increasingly used to analyze various human infectious disease dynamics such as malaria [11] , measles [12] , influenza [13] , leishmaniasis [14] and dengue [15] , [16] . We conducted wavelet analysis on a time series of reported HFRS cases to detect and to quantify variability of HFRS incidence over time . Wavelet analysis allows investigation and quantification of the temporal evolution of HFRS with different rhythmic components [9] . In this study , monthly HFRS incidences were square root transformed and normalized , and the trend was suppressed before analyses [9] . Bootstrap methods were used to quantify the statistical significance of the computed patterns [9] . As white noise or red noise hypothesis are not well-adapted to biological time series , we have chosen to use the synthetic series proposed by Rouyer et al . [17] , called “beta-surrogates” , which display a similar autocorrelation structure as the original time series . Using this approach , we thus obtain surrogates that mimic the shape of the original time series by displaying a power spectrum with the same slope in the log scale , but without exactly reproducing it [17] . This allows to test if the variability of the observed time series or the association between two time series is no different to that expected from a random process that has similar properties than the process underlying the observed time series . Wavelet time series analyses were applied using well-established algorithms [18] implemented in MATLAB software ( MathWorks , Inc . ) . A generalized linear model ( GLM ) with a Poisson distribution and a log link [19] was performed after adjustment for autocorrelation , seasonality , and lag effects . Potential seasonal variation was controlled by including the dummy variable month . To control the long-term trends , we created indicator variables for “year” of onset in the model . The variable “Moisture Condition” reflects the annual atmospheric moisture variability including annual precipitation and annual mean absolute humidity . The core model was: ( 4 ) where Var is the variance of the response and related to the mean μ , σ2 is the dispersion parameter . Our model for counts in the presence of overdispersion can be written as: ( 5 ) where p , q , r , s , u are lags determined by correlation analysis . η is the fitted model and Po is the Poisson distribution . Prec is the monthly accumulated precipitation and AH is the absolute humidity . MEI is the multivariate El Niño–Southern Oscillation index . Moisture Condition ( annual precipitation ) is calculated as sum of monthly precipitation for twelve months of the year and Moisture Condition ( annual mean absolute humidity ) is calculated by adding up all the monthly absolute humidity for twelve months of the year and then divide the sum by the number of months . A stepwise method was used to include variables , as long as there was significant improvement determined by calculation of the maximum likelihood ratio . GLM was fitted by using the R software package . A cross-correlation analysis was used to detect of the association between climate factors and HFRS transmission , with a different time lags . The two time series were filtered to convert to white noise before computing the cross-correlation .
A total of 9 , 130 cases were confirmed in Changsha between 1991 and 2010 , with a statistically significant gender difference . Individuals from lower socioeconomic groups ( students , workers and peasants ) comprised 93% of all HFRS cases , indicating these as high risk populations . The highest cumulative of annual incidence of HFRS during this period was 34 . 31/100 , 000 in 1997 , and the lowest 0 . 73/100 , 000 in 2008 . Figure 2 shows that HFRS cases in Changsha exhibit variability around two main temporal scales , with periods of 1 and 3–4 years . Both cyclic component were most pronounced in the 1990s . Timing of the breakpoint corresponds with apparent changes in both size of epidemics and patterns of seasonal variability , with appearance of the 1-year cycle in the wavelet spectrum . The wavelet spectrum also suggests that a 3–4-year period may have already been present prior to the 1990s . Clearly , there is a mark change in variability from the end of the 1990s to the beginning of the 2000s , accompanying a decline of overall incidence ( Figure 2 ) . Figure 3 shows that with increased annual precipitation , the annual incidence of HFRS cases began to increase from 1991 to 1995 . From 1996 to 1997 , annual precipitation increased quickly and reached a peak; the highest incidence of HFRS was recorded during the same period . With decreased annual precipitation , disease incidence began to decline . For example , annual precipitation began to decrease after 1998 , and annual incidence of HFRS also dropped sharply . In addition , the results reveal strong associations between annual HFRS incidence and annual precipitation ( Spearman rho = 0 . 667 , P<0 . 001 ) , and between annual incidence of HFRS and annual mean AH ( Spearman rho = 0 . 775 , P<0 . 001 ) . Disease incidence was associated with moisture over the long term . As shown in Table 1 , monthly HFRS cases are significantly correlated with precipitation , with the highest correlation coefficients having a lag of 5 months . The numbers of cases is correlated with AH and MEI , with a lag of 5 and 6 months , respectively . . Table 2 shows that the number of cases was third-order autoregressive , indicating that the number of cases in the current month was related to the number of cases in the previous 1 , 2 and 3 months . HFRS cases had association with AH , precipitation and MEI . Seasonal and long-term trends also contributed to the number of cases . The final model suggests the following: A 1 g/m3 increase in AH was associated with a 47 . 2% ( 95% CI , 42 . 3–52 . 3% ) increase in HFRS cases; a 1 mm increase in precipitation was associated with a 0 . 2% ( 95% CI , 0 . 1–0 . 3% ) increase in HFRS cases; a 1 mm increase in the moisture condition ( annual accumulated precipitation ) was associated with a 0 . 1% ( 95% CI , 0 . 1–0 . 2% ) increase in HFRS cases; a 1 g/m3 increase in the moisture condition ( annual mean AH ) was associated with a 2 . 6% ( 95% CI , 1 . 7–3 . 7% ) increase in HFRS cases; and a 1-unit MEI rise was associated with a 64 . 5% ( 95% CI , 43 . 5–88 . 5% ) increase in HFRS cases ( Table 2 ) . Only final parameter estimates of regression are presented in Table 2 . Finally , we compared the final model to a model with the autocorrelation , year , and month components alone , results of AIC and pseudo-R2 were shown in Table S1 . As shown in Figure 4 , the expected number of cases from the final model fitted very well with the observed number of cases in Changsha over the period 1991–2010 , including peak values; the pseudo-R2 value for the fitted model was 83 . 64% . In diagnosis of the residuals of the model , a random distribution was observed , with no autocorrelation among them .
Considering the dynamics of HFRs in Changsha , there was a large decrease of HFRS incidence after 1998–1999 . Possible explanations for this decrease could be due to the change in moisture conditions . We found a unique relationship between the moisture conditions and transmission of HFRS over a long period . The results most likely indicate that moisture not only influences growth of food sources that determine rodent population size , thereby affecting the HFRS transmission [20] , [21] , [22] , but also directly influences rodent activity and hantavirus infectivity [6] , [23] . The moisture condition ( annual precipitation and annual mean AH ) began to decrease after 1998 , and the disease incidence also began to decline during the same period . In our analyses , humidity and precipitation explained most the variance of the disease incidence . We postulate that because cultivated land and rivers cover most of the surface area in Changsha , these two landscape types can influence HFRS transmission by affecting the moisture conditions . Additionally , Changsha has been a traditional agricultural region where subtropical double-harvest rice has been historically cultivated , with at least 40% of the population as farmers or peasants . These combined factors support the transmission of HFRS . A key finding of this study was that atmospheric moisture conditions are an important predictor of the intensity of HFRS transmission in central China , the annual moisture condition can be a good indicator for predicting HFRS epidemic in the long run . We found a consistent association of the HFRS transmission with monthly precipitation , AH with 5-month lags , annual accumulated precipitation and annual mean AH . Based on moisture conditions over the long term , we improved the performance of the final model , with pseudo-R2 value of 83 . 64% . In the model we used AH rather than RH because it affects the abundance and distribution of rodents more directly , since AH is the actual water vapor content of air irrespective of temperature [24] . The MEI seems to precede oscillations of HFRS in Changsha . This may reflect the timing of relationships between El Niño and local climate . El Niño can affect large-scale phenomena , which generate local climatic phenomena and thereby influence oscillations of epidemics . The El Niño effect may act as a pacemaker to affect local climatic moisture conditions , which in turn affect HFRS incidence . Higher precipitation levels can induce severe HFRS epidemics in Changsha; however , excessive precipitation may cause flooding and a reduction of rodent population . The complexity of the links between HFRS dynamics and moisture is emphasized by the significant correlations in the annual mode over a long term observed in this study . In summer 1998 , there was a large-scale weather anomaly in China , and the entire Yangtze River basin ( including Changsha in the Xiangjiang River basin ) suffered a severe flood event , the largest during the 20th century apart from the 1954 floods . We speculate that a reduction in transmission was caused by declining rodent populations , reducing their contact with humans . As Thibault and Brown showed , an all-time record low number of rodents were captured the month following the downpour and resulting flooding [25] . The limitations of this study should also be acknowledged . Many factors can contribute to the transmission of HFRS . In addition to changes in climate conditions , the obvious decrease in the 2000s could also be due to other factors , e . g . changes in socio-economic status , human activities and movement , population immunity and/or local and national intervention programmes . The findings here show that atmospheric moisture is an important predictor of HFRS incidence . This work may contribute to characterizing the temporal dynamics of HFRS in China , or in other countries with similar regional climate conditions . It may also have significant implications for integrating climate monitoring and disease surveillance data to effectively control and prevent the epidemics of HFRS . | Hemorrhagic fever with renal syndrome ( HFRS ) , a rodentborne disease caused by Hantaviruses , is characterized by fever , haemorrhage , headache , back pain , abdominal pain , and acute kidney injury . At present , it is endemic in all 31 provinces , autonomous regions , and metropolitan areas in mainland China where human cases account for 90% of the total global cases . Infection rates and population dynamics of hosts are thought to be influenced by climatic factors , especially humidity . Some studies have found that hantaviruses are limited in their spread to high-humidity environments for extended ex vivo stability . Here we provide the evidence that HFRS incidence was strongly associated with moisture conditions , including seasonal variation and annual situation , in Changsha , mainland China , 1991–2010 . The results most likely indicate that moisture not only influences growth of food sources that determine rodent population size , thereby affecting the HFRS transmission , but also directly influences rodent activity and hantavirus infectivity . These findings offer insights in understanding possible causes of HFRS transmission , and can be used in disease surveillance and risk management to provide early warning of potential epidemics of this disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"environmental",
"health",
"zoonoses",
"hantavirus",
"environmental",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"modeling",
"public",
"health"
] | 2013 | Atmospheric Moisture Variability and Transmission of Hemorrhagic Fever with Renal Syndrome in Changsha City, Mainland China, 1991–2010 |
The multiple-breath washout ( MBW ) is a lung function test that measures the degree of ventilation inhomogeneity ( VI ) . The test is used to identify small airway impairment in patients with lung diseases like cystic fibrosis . However , the physical and physiological factors that influence the test outcomes and differentiate health from disease are not well understood . Computational models have been used to better understand the interaction between anatomical structure and physiological properties of the lung , but none of them has dealt in depth with the tracer gas washout test in a whole . Thus , our aim was to create a lung model that simulates the entire MBW and investigate the role of lung morphology and tissue mechanics on the tracer gas washout procedure . To this end , we developed a multi-scale lung model to simulate the inert gas transport in airways of all size . We then applied systematically different modifications to geometrical and mechanical properties of the lung model ( compliance , residual airway volume and flow resistance ) which have been associated with VI . The modifications were applied to distinct parts of the model , and their effects on the gas distribution within the lung and on the gas concentration profile were assessed . We found that variability in compliance and residual volume of the airways , as well as the spatial distribution of this variability in the lung had a direct influence on gas distribution among airways and on the MBW pattern ( washout duration , characteristic concentration profile during each expiration ) , while the effects of variable flow resistance were negligible . Based on these findings , it is possible to classify different types of inhomogeneities in the lung and relate them to specific features of the MBW pattern , which builds the basis for a more detailed association of lung function and structure .
The multiple-breath washout ( MBW ) is a lung function test that measures the degree of ventilation inhomogeneity [1] and is increasingly used for both research and clinical purposes in patients with obstructive lung disease , such as cystic fibrosis , primary ciliary dyskinesia , etc . [2–4] . The test is based on the clearance of a tracer gas during multiple tidal breaths . Each MBW test comprises a washin and a washout phase . During the washin phase , the tracer gas ( normally an inert extrinsic gas ) is delivered in a known concentration . When the tracer gas concentration reaches an equilibrium in the lung , the washout phase starts . In the case of inert intrinsic gases ( N2 ) , the test is simplified , as no washin phase is needed . In the washout phase the subject inhales a gas other than the tracer gas ( e . g . pure O2 in case of N2MBW ) , so that the lungs wash out the tracer gas gradually by each expiration [2] . The progressive decrease in tracer gas concentration during the washout ( washout envelope ) as well as the breath-by-breath analysis provide useful information about the distribution of ventilation within the lung [1] , and for this reason the test is of increasing importance for the medical community . However , the biomechanical phenomena that influence the specific washout profile of a tracer gas are not well understood . The anatomical diversity in the airways as well as the physiological properties of the lung tissue ( e . g . compliance , resistance ) influence the respiratory function in a complex way [5 , 6] . Over the last years , computational models have been used to better understand those interactions . Anatomically based three-dimensional ( 3D ) lung models have elegantly simulated lung tissue mechanics [7–9] and 3D fluid dynamics in large airways [10–12] , but without specific focus on the ventilation and gas transport in the entire lung in the context of MBW . More simplified models have addressed these phenomena , but either have not included physiological asymmetries in lung morphology , or have not modeled the whole lung [13–23] . Therefore , results of these models cannot be directly compared to clinical data in humans [24] . The aim of this article is to introduce a computational model that simulates gas transport in the entire lung during the MBW test , taking into account transport phenomena at different scales . It should allow relating physiological phenomena at the smallest scales of the lung to gas concentrations at the mouth , which can be measured clinically . Such simulations can provide detailed insight in the gas transport dynamics during the MBW in different airways and help to understand better the effect of lung morphology and tissue mechanics on tracer gas washout . To this end , structural data from human lungs have been used to construct a fractal lung model , taking into account morphological and physiological asymmetries in lung anatomy [15 , 25 , 26] including the number and size distribution of the acini [27 , 28] . We have designed the model in a way that is complex enough for the introduction of different types of asymmetries and structural properties of the acini , but simple enough to allow for quick computational turn-around times on desktop computers . The model was used to study some basic mechanisms , which lead to well-known clinical observations in MBW [1] , and an example of a model for the healthy lungs was created , using known parameters that can produce a physiological heterogeneous ventilation distribution , as described in healthy individuals [6] . Such models inherently lack suitable validation methods [29] because ground truth data is not available in the lower airways . Here , we compared the MBW simulations in the computational lung model to data from nitrogen MBW ( N2MBW ) tests from healthy controls ( N = 4 ) . Although this comparison is too small to serve as a statistically solid validation of the model , the good agreement between computational results and in vivo data illustrates the potential of the proposed model to reproduce clinical test data .
The study was approved by the Ethics Committee of the Canton of Bern , Switzerland ( KEK-Gesuchs-Nr: 181/03 ) , and caregivers gave written informed consent . In our model , the airway morphology is represented by a generic dichotomous tree network of straight branching pipes [30] , which terminates in trumpet-like compartments . The straight pipes represent the larger , non-compliant airways , where convective transport is dominant . The dimensions of the trachea and the relative dimensions ( i . e . with respect to the trachea ) of the airways in the first four generations are defined according to anatomical data from Weibel [26] . In addition , the entire airway network of pipes ( all conducting airways including the trachea , without the trumpet-like compartments ) was then scaled to meet a specific functional residual capacity ( FRC ) that is , the air volume that remains in the lung after tidal expiration . The length of the scaled trachea was defined as l0=l0W ( FRCFRCW ) 1/3 ( Eq 1 ) where the subscript W indicates a reference quantity [26] . For airways past the 4th generation , we applied the scheme for a regular branching asymmetry introduced by Majumdar et al . [25 , 30] . In this scheme , each pipe-like airway bifurcates in a major and a minor daughter airway . In the airway network every parent , minor , and major daughter share a common node . The dimensions of the daughter airways , namely their diameter and length , are different fractions of the dimensions of their common parent pipe , dz+1maj=dzκmajwithκmaj= ( 1−r ) 1/ηanddz+1min=dzκminwithκmin=r1/η ( Eq 2 ) where dz is the diameter of a pipe at generation z , and r and η denote the asymmetry parameter and reduction rate , respectively . The same scheme was used to define the airway lengths . Majumdar et al . proposed values r = 0 . 326 and η = 2 . 97 for which the resulting structure best represents the human lung on a statistical basis [25] . In the present model , this bifurcation scheme was applied recursively until the diameter of a pipe was less than a lower limit diameter dlim = 1 . 8 mm . This limit defines the terminal pipes in the model and was chosen to represent bronchi at the 8th-12th generation . The generation of the individual terminal pipe-like airways depends on the limit diameter and varies within the model , due to the asymmetric bifurcation scheme . This limit diameter was chosen in order to complete the simulations with reasonable computational costs . A numerical experiment showed very small differences in the simulation outcomes for dlim = 1 . 6 , 1 . 8 , 2 . 0 mm . The computational unit distal to a terminal pipe constitutes a lobule and was modelled using a trumpet-like compartment ( trumpet lobule ) . Please note that the “lobule” as defined here differs from the anatomical term lobule as defined by Miller [31] , because it contains , apart from the acini that include respiratory bronchioles , alveolar ducts and alveoli , also the generations of conducting airways with diameter smaller than dlim . Unlike the pipes , the trumpet lobule is a compartment with diverging , time-variable cross-section ( Fig 1A ) . The number of these trumpet lobules is equal to the number of terminal pipes , as each terminal pipe leads to one lobule . The residual volume of the trumpet lobules was defined such that the total volume of the model ( pipes and trumpet lobules ) at the end of a tidal expiration equaled a predefined FRC . Fig 1B shows a sketch of a model lung illustrating the relative scales of a network of airways defined by the bifurcation rule ( Eq 2 ) . For the trumpet model representing the lobules and their peripheral airways , a model for the total cross-section of the trumpet lobule Slb ( x , t ) , as well as for the mean advection velocity ulb ( x , t ) had to be derived . The flow rate at the inlet of each trumpet equals the flow rate in a terminal pipe and Qt ( t ) was known from the results of a model for lung ventilation ( described in the upcoming section ) and was used to compute the total volume of the trumpet Vlb ( t ) =∫0tQtdt+Vlb0 . The initial volume of the trumpet lobule , Vlb ( t = 0 ) = Vlb0 , followed from FRC-based scaling of the lung model . Assuming a uniform homothety ratio of κ = 0 . 85 [19] for airways lumped in a trumpet lobule , the total change of cross-section along the streamwise coordinate x can be described as S ( x , 0 ) =Stκ^z ( x ) , withκ^=2κ2 ( Eq 3 ) where St is the cross-section of the terminal pipe , and z ( x ) is the generation at position x with respect to the inlet of the trumpet lobule where z = x = 0 . Considering lt to be the length of the terminal pipe , the cumulative length at generation z ( with respect to the inlet of the lobule ) would be ∑k=1zltκk . The limits of this sum for z→∞ are limz→∞∑k=1zltκk=ltκ/ ( κ−1 ) ) ≕L . ( Eq 4 ) From these relations , an expression for the generation in function of the distance to the inlet of the trumpet can be computed , z ( x ) =log[xκ−1κlt+1]log ( κ ) ( Eq 5 ) Using Eq 3 together with Eq 4 for further treatment of the lobule model becomes a rather cumbersome task . We therefore sought a model Slb ( x , t ) , which approximates Eq 5 but allows to derive an analytical expression for ulb ( x , t ) . To this end , we used a power law of the form Slb ( x , t ) =p1xn1+p2xn2+St , ( Eq 6 ) with n1 , n2 = 20 , 2 , where the coefficients p1 and p2 were defined such that the lobule cross-section Slb intersects with Eq 3 at a chosen generation z* , and the prescribed initial volume Vlb0 of the trumpet lobule is obtained for a given length llb of the trumpet lobule . The mathematical expression for this parameter definition as well as a graphic comparison between the formulation Eq 3 and the model Eq 6 are provided in S1 Appendix ( Section 1 ) . The expression given in Eq 6 determines the shape of the trumpet lobule at any time . An important feature of this model is the major contribution of peripheral airway ( where x is close to llb ) to the overall expansion of the lung . Furthermore , the differentiation with respect to time of Eq 6 as well as the integration along the trumpet centre line coordinate x is straightforward and therefore allows to formulate a modified advection diffusion equation for the trumpet lobule ( see S1 Appendix , Section 1 ) Apart from the constraint on the total volume , the geometrical and mechanical properties of the stiff ( pipes ) and the compliant parts ( trumpet lobules ) can be modified individually to study systematically the effects of structural lung inhomogeneity . Lung ventilation was simulated by a lumped parameter ( 0-dimensional ) model ( Fig 2 ) based on the model morphology described above . For the pipe-like airways , purely resistive elements were used . Womersley's theory for pulsatile flow in tubes [32] was applied to account for inertial effects during normal breathing , which have a considerable effect on the flow resistance in bigger airways until the fifth generation [33] ( see also S1 Appendix ) . In general , the pressure difference between two subsequent nodes with index i and j in the network of pipe-like airways reads pi−pj = RijQij . Here Qij is the flow rate from node i to node j , and the hydrodynamic resistance Rij depends on the radius rij and the length lij of the conducting airway between two nodes , and on the breath period TB . More information on the pressure-flow relation can be found in S1 Appendix . The trumpet lobule model , mainly representing compliant airways , is composed of a nonlinear compliance element ( elastic pressure , pel ) and a resistance element ( viscous pressure loss , pdiss ) , acting in series between a node i , corresponding to a terminal pipe , and the pleural gap with pressure ppl . This representation is based on the pressure-volume relation as presented by Bates [34] . However , instead of the regular linear elastic law , we used an empirical exponential expression to represent pel . The corresponding pressure difference was defined as pi−ppl=βeγVlb0 ( eγV˜lb−1 ) ⏟pel+RlbQlb⏟pdiss ( Eq 7 ) where Rlb , Vlb ( t ) =Vlb0+V˜lb ( t ) and Qlb denote the total flow resistance of the lobule , its volume and the flow rate into the trumpet lobule , respectively; Vlb0 is the lobule volume at FRC and V˜lb ( t ) is the dynamic volume during breathing [34] . The shape parameters β and γ were used to define and modify the degree of the compliance parameter φ ( i . e . the distensibility of the lobule , described below in details ) and the non-linearity ( in the pressure-volume relation ) of the trumpet lobule , respectively . There is a single β-γ pair for each lobule , and both are defined indirectly by fixed intersection points in the pressure-volume curve for the elastic pressure ( see Fig 3A ) . More information on the mechanical properties of the model can be found in the Section “Structural and mechanical modifications of the model” . To ensure mass conservation , the flow rates in and out of each node i satisfy the balance ∑Q = 0 . At the inlet , i . e . upstream node of the trachea , the flow rate is defined by the predefined boundary conditions Qin ( t ) , and for the trumpet lobule the flow rate of the terminal pipe Qt equals the change of volume of the lobule , Qt=dVlbdt . ( Eq 8 ) This results in a system of differential algebraic equations ( DAE ) . Together with appropriate boundary conditions , the DAE system governs the flow distribution in the lumped parameter model . A detailed description of the numerical implementation of the lumped parameter model is provided in S1 Appendix . During inspiration , gas is transported from the mouth and the nose toward the alveolar membrane , as a result of the motion of the diaphragm and the thoracic cavity , causing a volume increase in the lung and a pressure decrease in the pleural gap . This negative ( relative ) pressure in the pleural gap causes a pressure gradient across the peripheral lung tissue and along the airways to the mouth and nose . Therefore , the pleural pressure and the pressure at the mouth would be obvious choices for the boundary conditions for the lumped parameter model ( LPM ) . However , the pleural pressure is in general not measurable during clinical routine . Instead , the pleural pressure was determined computationally such that a prescribed flow rate Qin ( t ) was attained in the trachea . This flow rate could be easily determined during clinical testing . Gas transport in the lung occurs by convection and diffusion [35 , 36] . To model the distribution of MBW tracer gases in the lung , typically N2 or sulfur-hexafluoride [1] , a one-dimensional advection-diffusion transport equation for a scalar variable representing the normalized gas concentration c = c ( x , t ) ∈[0 , 1] was solved with a finite difference method ( detailed description in S1 Appendix ) . Simulation of MBW involved the computation of gas concentration in all model airways over multiple breath periods . We considered only inert gases for which the dominant transport mechanisms are advection by the carrier gas ( ambient air ) and diffusion . The transport process was modelled along the centerlines of each pipe-like airway and along the axis of each lobule trumpet . The 1D approximation entails consideration of averaged ( lumped ) quantities within a cross-section , or in case of the trumpet model , within all airways represented by the cross-section of the trumpet at a given axial position . Advection-diffusion processes in a pipe-like geometry are subject to high radial velocity gradients , which can lead to radial concentration gradients such that the average effective diffusivity within a cross-section increases . In the human lung , these phenomena can take place at different scales: In the upper airway , where the Reynolds number Red ( based on the airway diameter d ) is about 10'000 , turbulent flow strongly enhances mixing . In smaller airways ( Re≪2000 ) , Poiseuille flow with a parabolic velocity profile can be assumed . Enhanced diffusion due to high velocity gradients was modelled based on the concept of Taylor dispersion , where the local diffusion coefficient is a function of the local Peclet number [37] , D^=D ( 1+1192Pe2 ) ( Eq 9 ) where D^ and D are the effective and molecular diffusion coefficients , respectively , and Pe = ud/D is the Peclet number defined with the local mean velocity u and the local airway diameter d . This is an approximate measure . Taylor dispersion is a concept , which strictly applies only to developed flow in straight tubes . Hence , the modification of the diffusion coefficient ( Eq 9 ) does not account for local changes in diffusivity due to secondary flow phenomena occurring at bifurcations and curved airways , and likely underestimates the level of diffusion enhancement . The advection-diffusion equation for gas transport was solved separately in each airway , applying interface conditions at the bifurcation nodes to couple the transport between different airways . We considered a transport equation of the following general form ∂ ( Sc ) ∂t+∂F∂x=0 ( Eq 10 ) with the flux F=Saduc−SD^∂c∂x ( Eq 11 ) Here , we distinguished between the total cross-section S and the cross-section Sad in which advection with the carrier gas velocity u ( x , t ) takes place . This assumption is important , because the airway geometry becomes increasingly complex ( i . e . non-tubular , alveolar ducts and alveolar trees ) for smaller airways , and tracer gas advection does not occur in the entire lumen [35 , 38 , 39] . At bifurcations , the concentration flux was conserved by enforcing ∑F = 0 . In S1 Appendix , a model for the trumpet lobule is derived which allows a more specific form of Eq 10 to be stated for the gas transport within trumpet lobules with non-constant cross-section . In addition , the finite difference scheme for spatial derivatives and time-integration method used for the numerical solution of Eq 10 are explained . The proposed multi-scale model was designed to study the effects of functional and structural inhomogeneities of the peripheral airways on the N2 washout procedure . To illustrate the capabilities of the model , several parameters of the trumpet lobules were systematically modified . The parameters were defined in a way that their effect on lung geometry and mechanics was physically meaningful and intuitively clear: The aim of the comparison of model results with in vivo data was to demonstrate that the model can relate microscale modifications at the lobular level to clinically observed MBW metrics ( MBW washout envelope and phase III slope analysis [1] ) . We used data from healthy adolescents recruited for lung function studies in the Inselspital Children’s University Hospital , Bern , Switzerland . N2MBW measurements ( N = 4 ) were collected according to the recent ERS/ATS consensus guidelines [1] using the ultrasonic flowmeter ( Exhalyzer D , Eco Medics AG , Duernten , Switzerland ) and the software provided by the manufacturer ( Spiroware 3 . 1 . 6 ) as previously described [49] . During the test , the subject sat in an upright position wearing a nose-clip and was asked to breathe regularly through a snorkel-like mouthpiece connected to a bacterial filter and a dead space reducer .
A baseline simulation has been performed with the parameter settings listed in Table 1 , but without any added asymmetries with respect to lobular compliance , lobular residual size , or lobular resistance . We compared this reference simulation with results for modified lobular compliance where two regions ( each accounting for 25% of all trumpet lobules ) were modified using ϕ = 0 . 5 and ϕ = 1 . 5 , respectively . Fig 4 shows the results for both simulations ( baseline and compliance modifications ) . The washout curve ( Fig 4C and 4D ) can be analyzed in different ways: Each expiration starts with zero N2 concentration ( phase I ) , which corresponds to the washout of the dead space ( the pipes in the model ) , and then a rise in N2 concentration , which is first very quick ( phase II ) and later slow ( phase III ) [50–52] . Phase III is the part of the concentration-expired volume curve that corresponds to 50% to 95% of the expired volume per breath [1] ( Fig 5 ) . The end-expiratory N2 concentration diminishes from one breath to the next as the N2 washout progresses . The ratio of two subsequent end-expiratory N2 concentration values , i . e . the decay of the washout curve envelope , provides information about the gas washout efficiency of the whole lung , and indirectly about the ventilation of lung compartments . For the baseline model configuration , this exponential decay is uniform ( linear envelope graph in a semi-logarithmic plot over time , in Fig 4D ) . In the model with regions of variable lobular compliance , the decay is non-uniform and delays towards the end . This can be associated with a non-uniform ventilation in different regions of the lung . To quantify these washout properties , an exponential function of the following form was fitted to the envelope of end-expiratory values . f ( t ) =Ae−α1t+ ( 1−A ) e−α2t ( Eq 13 ) Here , α1 and α2 ( with α1 , α2>0 ) are the decay rates of two superimposed processes , one fast and one slow [53] , and A∈[0 , 1] represents the relative weight of the slow decay process ( α1<α2 ) . The decay of the process is considered uniform if A = 1 and non-uniform if A<1 . Changing the compliance properties as described above resulted in A = 0 . 44 , α1 = 0 . 018 , and α2 = 0 . 045 , compared to A = 1 . 0 and α1 = 0 . 031 for the baseline . The spatial distribution within the lung model after the fifth breath ( Fig 4E and 4F ) provides further insight to the non-uniform ( i . e . heterogeneous ) washout process . The concentration remains high in lobules with decreased compliance ( ϕ = 0 . 5 ) and is reduced more rapidly in those with increased compliance ( ϕ = 1 . 5 ) . Later in the washout process , the relative concentration difference between the two regions is therefore higher than in the baseline configuration and the washout slows down due to the increasingly dominant contributions of the slowly washed-out units . The concentration profile can be further analyzed and interpreted on a per-breath basis . The slope of the concentration-volume curve during phase III ( slope III ) indicates whether the gas mixtures from different lung regions have different gas concentrations [35 , 50] ( Fig 5 ) . Therefore , slope III has been linked to inhomogeneous gas transport dynamics , and parameters derived from this analysis are of increasing clinical relevance [54–56] . To quantify the phase III profile in our study , the parameter sIII is defined , which is the slope of a linear function fitted to the N2 concentration values corresponding to the phase III , normalized with the mean N2 concentration during phase III ( Fig 5 ) : SIII=SlopeIIImeanN2 ( phaseIII ) ( Eq 14 ) Comparing the baseline washout with the washout of modified compliance , sIII was higher ( steeper slope ) in the modified lung already for the first breath ( Fig 4A ) , and this difference in sIII increased further until the last breath ( Fig 4B ) . For the comparison of other types of lung model modifications that represent a specific kind of structural and functional lung asymmetry , we used the weight parameter A and the decay rate α1 ( and α2 ) from the fitting function ( Eq 9 ) approximating the washout envelope , as well as the clinically important parameter sIII ( Eq 14 ) . In separate simulations , the residual volume of trumpet lobules was altered using θ = 0 . 5 for a subset of lobules accounting for 25% of all trumpet lobules . Note that the model is designed such that it scales the residual volume of the remaining 75% of lobules automatically to preserve the prescribed FRC of the lung . In another simulation , the resistance was increased ( τ = 8 ) in 25% of the lobules , which corresponds to a local reduction of airway diameters by approximately 40% . Fig 6 shows the simulation results for these three modified cases together with the baseline configuration . Note that the baseline results ( black line in Fig 6 ) are barely visible , because some of the simulations in the model with increased resistance yielded very similar results . The compliance modifications showed the strongest effect and yielded a non-uniform , delayed washout ( Fig 6C and 6D ) . While the resistance modification did not lead to any notable difference from the baseline result ( A = 1 . 0 ) , heterogeneous residual lobular volume also yielded a non-uniform washout ( A = 0 . 84 ) similar but weaker than the model with modified compliance . Interestingly , the washout was delayed in later breaths in the case of modified residual lobular volume , and of compliance modifications . The modified residual volume caused in addition a faster decay for early breaths ( Fig 6C before t = 60 s ) . This was also reflected in the slow ( α1 = 0 . 028 ) and fast ( α2 = 0 . 074 ) decay rates , e . g . when compared to the decay rate for the baseline configuration ( α1 = 0 . 031 ) . A reason for this could be that in smaller and shorter lobules , the flow of pure oxygen replaces a larger fraction of N2 and thus the washout per breath is more efficient ( faster ) . Although the residual lobule size is different , a similar flow rate of pure oxygen is preserved in all lobules as long as the compliance properties and the pleural pressure distribution remain uniform . The comparison of normalized phase III slopes sIII also clearly discriminated the compliance modifications from the other cases . In a per breath analysis , a flat plateau-like profile resulted for the baseline , as well as for the cases with altered lobule size and resistance . For regionally altered lobule compliance , however , the gas concentration increased nearly linearly in phase III . This difference is already visible from the first breath , but becomes more prominent as the washout progresses ( Fig 6A and 6B ) . For lobular resistance and volume modifications , the phase III concentration profile remained a flat plateau during the whole washout . In the case of modified lobule compliance sIII increased from breath to breath . These different trends in sIII are illustrated in Fig 7 . Increased slopes sIII indicate that concentration differences at airway bifurcations are increasing in phase III , such that contribution from high concentration units becomes more and more dominant . This was only the case for compliance modifications , where unequal rates of pure oxygen feed into different lobules . In case of modified volume differences , the spatial concentration distribution was not uniform , but did not change during phase III . In the next section , the relation between spatial concentration distribution in the lung and the phase III concentration profile is further discussed . In summary , the compliance modifications had the strongest impact on gas washout in terms of non-uniform concentration decay and sIII values . A heterogeneity in residual lobule volume has a notable influence on the washout profile during early and later breaths , while the phase III concentration profile does not differ from the baseline . Compared to these two cases , the changes in hydrodynamic resistance have a negligible effect on gas washout profile . This is in accordance to a previous report [7] . To study the capabilities of the model in reflecting the effect of different spatial distributions of structural inhomogeneities , we look again at the example of compliance modifications . Instead of applying these modifications to a subset of neighboring lobules ( regional distribution ) , we distributed the modifications over 50% of the lobules regularly distributed in the entire lung domain ( local distribution ) . This distribution intends to mimic lung diseases such as cystic fibrosis that do not spread in a locally organized manner [57 , 58] . To this end , we performed simulations where for every other lobule the compliance was alternately reduced ( ϕ = 0 . 5 ) or increased ( ϕ = 1 . 5 ) . Local and regional compliance modifications yielded similar results for the whole washout ( Fig 8 ) . Both caused a non-uniform decay with A = 0 . 52 ( local ) and A = 0 . 44 ( regional ) . Minor differences were found in the slow and fast decay rates ( α1 = 0 . 020 vs . 0 . 018 and α2 = 0 . 044 vs . 0 . 045 ( local vs . regional ) ) . However , the washout curve differed considerably with respect to phase III slope sIII ( Fig 8A and 8B ) . Starting with only small differences in the first breath , sIII increased with every breath for regional modifications , while it decreased for local compliance modifications , reaching slightly negative values after approximately 35 simulated breaths ( Fig 9D ) . The spatial concentration distribution at the end of the fifth breath is depicted for both local and regional compliance modifications in Fig 9A and 9B . The lobular concentrations were clearly different for lobules with altered compliance properties . Trumpet lobules with increased compliance properties showed higher ( normalized ) N2 concentration and vice versa for reduced compliance . Furthermore , the comparison of the concentration distributions for regional and local compliance heterogeneity suggest the following mechanisms governing the phase III slopes: 1 ) The mixing of gas due to concentration gradients at airway bifurcations occurred at different levels: For regional modifications , the mixing took place at several generations of large airways , starting from the main bronchi to smaller conducting airways , where gas from different regions comes together , causing an increasingly non-uniform concentration distribution within the network . For local modifications , the mixing took place at already small airways and only over a few generations immediately above the lobules , which leads to a more uniform concentration distribution in the large airways . 2 ) Spatial concentration differences throughout the airway tree increased over time for regional modifications , while they remained approximately constant for local modifications ( see S1 Video and S2 Video ) . This may be an explanation for the increasingly different phase III slopes ( Fig 9C and 9D ) . In order to mimic the small degree of inhomogeneous ventilation that exists physiologically in the healthy lungs [6] , we used a combination of regional modifications of lobular size ( θ = 0 . 5 in two regions each accounting for 12 . 5% of all trumpet lobules ) and lobular compliance ( ϕ = 0 . 5 , ϕ = 1 . 5 in two regions respectively , each accounting for 12 . 5% of all trumpet lobules ) to approximate normal structural and mechanical inhomogeneities of the lung . The modifications parameters were determined through an iterative process: First , the total set of lobules was split in sub-regions ( each 12 . 5% ) accounting for modified lobular compliance and size . From the previously conducted modification parameter study it was known that a region with smaller lobule volume ( θ<1 ) would account for faster washout during early breaths , while a region of less compliant lobule ( ϕ<1 ) would cause the washout to be delayed ( less efficient ) towards later breaths . In a first step , the modification parameter θ was altered until the washout envelope of the measured and the simulated N2MBW curve were in agreement for the first 5–10 breaths ( qualitatively , by visual inspection ) . Subsequently , the modification parameter ϕ was altered , which did not change the results from the first fitting step , until the washout envelopes also approximately matched for the last 30–50 breaths . In a MBW test simulation using this model , the lung clearance index ( LCI ) [1] was with 6 . 1 in the normal range reported in the literature [56] . S3 Video ( supporting material ) visualizes the evolution of the spatial concentration distribution over 50 breaths in such a model for a sinusoidal inlet flow profile . It shows the asymmetric concentration distribution due to regional differences in lobule size and lobule compliance , and how these differences evolve over time . Demographics about the healthy subjects and lung volumes for each test simulation are provided in Table 2 . For each simulation , we used the model for the healthy lung and applied the flow rate profile and the FRC as measured during the MBW test . The FRC was used to scale each lung model , using Eq 1 with the reference value FRCW = 3 . 5 l [26] . Fig 10 and Table 3 show measured concentration curves and MBW outcomes , respectively , for N2MBW measured in healthy subjects together with corresponding simulations . Overall , the simulated washout envelopes for the four cases were in good agreement with the experimental data . On a breath-per-breath basis , differences were more prominent . For example , end-expiratory concentration values were moderately different for several breaths . A higher variability was found during phase III . Possible reasons for this could be: the complex structural and mechanical asymmetries in the healthy lung , which are not sufficiently modelled with the types of inhomogeneities used in this study ( lobular compliance , residual size , and resistance ) ; the non-uniform breathing pattern of the healthy subjects; and measurement errors . A detailed explanation or justification of these discrepancies was beyond the scope of this study . The presented model has several limitations . For the sake of simplicity , several physiological and anatomical features were either idealized or neglected . The trumpet lobules include the last generations of the conducting airways , in order to keep the computational costs reasonable . We acknowledge that the acinar geometric properties are not entirely simulated in the lobule and that parameters like the homothety ratio are , in reality , not uniform throughout the airways contained in the lobule . It was beyond the scope of this study to investigate the effects of different cross-sectional development in the acinar region on the MBW washout curve . Although the dimensions of the model ( airway diameter and length per generation , FRC , lobular size distribution ) were derived from anatomical data , the overall geometry of the model is rather generic and not anatomically based . Spatial asymmetries between the right and the left lung were not modelled . The FRC-dependent scaling of the modeled airway tree allows using it for MBW simulations in a large FRC range , however anatomical differences between children and adults were not introduced in the model . Pressure losses due to turbulent flow in the big airways and flow changes at the bifurcations and/or the heavily curved airways are not included in the model , as these phenomena require 3D anatomical data . However , the results in this and in previous studies [7] demonstrate that flow resistance plays a smaller role in ventilation inhomogeneity compared to differences related to the loading and constitution of respiratory units . Although there is a spatial pleural pressure gradient in real lungs , the pleural pressure was assumed to be spatially uniform , because , to our best knowledge , there is no realistic estimate for this gradient . We further assumed that the initial mechanical stress distribution is uniform between the lobules although variability should be expected in real lungs . It will be interesting to investigate the effect of this variability and effects due to gravity on future studies . Next , regarding the model for the healthy lungs , we acknowledge that various other combinations of θ and ϕ , as well as introduction of other modifying parameters may lead to a similar MBW profile and outcomes . However , our aim was neither to mimic breathing characteristics on the individual level [59] , nor to relate MBW outcomes to particular local anatomical characteristics , but to provide an example that resembles healthy lungs , based on principles of lung physiology . Finally , no specific model for the upper airways , the mouth , and the interface equipment has been used in the simulations . The additional pathway for gas transport introduced by these parts was represented by an elongation of the straight pipe representing the trachea . We presented a multi-scale model of the whole lung that simulates the gas transport and washout in conducting and acinar airways , including non-linear tissue mechanics . In order to mimic a physiological degree of ventilation inhomogeneity as described in healthy lungs , we introduced modifications in mechanical and geometrical properties on a lobular level . This study demonstrates that regional and local alterations of airway properties have different effects on the expiratory phase III in the MBW . Phase III slope profiles were notably more pronounced and sensitive to the degree of modifications for regional type modifications compared with local type modifications . Furthermore , the study revealed the different functional relations between the MBW concentration curves and airway compliance , volume and flow resistance . Increased heterogeneity of lobular compliance and residual volume correlated with a delayed washout , while heterogeneous flow resistance had a negligible impact . Finally , the simulation results are in accordance with real MBW data obtained from healthy subjects , on a qualitative level . The model can be used to study MBW characteristics in health and disease . It offers the opportunity to understand the ventilation distribution in the healthy lung , and to investigate more profoundly MBW features that extract localized information , like the slope III analysis . By applying modifications in mechanical properties that exceed the physiological limits , the model can also mimic certain patterns of lung disease . Thus , it can be used to study the effect of such diseases on MBW concentration curves . In addition , the model may also serve as a tool to visualize gas transport in the lung during a MBW test , which could support patient education . | Obstructive lung diseases , like cystic fibrosis or primary ciliary dyskinesia , lead to inhomogeneous ventilation . The degree of observed inhomogeneity represents a clinical measure for the progression of the disease . The multiple-breath washout ( MBW ) is a lung function test that measures this inhomogeneity in the lung . However , the factors that influence the results of the test and differentiate between health and disease are not well understood . Computational models help us to understand better the relation between anatomical structure and physiological properties of the lung , but none of them has dealt in depth with the MBW test in whole . Our aim was to create a lung model that simulates the entire MBW test and study the role of lung structure and tissue mechanics on the washout procedure . We developed a multi-scale lung model to simulate the inert gas transport in all airways including the gas exchange area . Our model offers the opportunity to understand the ventilation distribution in the healthy lung . It can also mimic certain patterns of lung disease by applying modifications in mechanical properties out of the physiological limits . Thus , it can be used to study MBW characteristics in health and disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"&",
"discussion"
] | [
"tissue",
"mechanics",
"medicine",
"and",
"health",
"sciences",
"classical",
"mechanics",
"fluid",
"mechanics",
"biomechanics",
"pulmonology",
"pleurae",
"simulation",
"and",
"modeling",
"respiratory",
"system",
"trachea",
"research",
"and",
"analysis",
"methods",
"respiratory",
"physiology",
"fluid",
"dynamics",
"pulmonary",
"function",
"biophysics",
"continuum",
"mechanics",
"fluid",
"flow",
"physics",
"flow",
"rate",
"anatomy",
"physiology",
"thorax",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences"
] | 2019 | A multi-scale model of gas transport in the lung to study heterogeneous lung ventilation during the multiple-breath washout test |
Recognition of specific cell surface glycans , mediated by the VP8* domain of the spike protein VP4 , is the essential first step in rotavirus ( RV ) infection . Due to lack of direct structural information of virus-ligand interactions , the molecular basis of ligand-controlled host ranges of the major human RVs ( P[8] and P[4] ) in P[II] genogroup remains unknown . Here , through characterization of a minor P[II] RV ( P[19] ) that can infect both animals ( pigs ) and humans , we made an important advance to fill this knowledge gap by solving the crystal structures of the P[19] VP8* in complex with its ligands . Our data showed that P[19] RVs use a novel binding site that differs from the known ones of other genotypes/genogroups . This binding site is capable of interacting with two types of glycans , the mucin core and type 1 histo-blood group antigens ( HBGAs ) with a common GlcNAc as the central binding saccharide . The binding site is apparently shared by other P[II] RVs and possibly two genotypes ( P[10] and P[12] ) in P[I] as shown by their highly conserved GlcNAc-interacting residues . These data provide strong evidence of evolutionary connections among these human and animal RVs , pointing to a common ancestor in P[I] with a possible animal host origin . While the binding properties to GlcNAc-containing saccharides are maintained , changes in binding to additional residues , such as those in the polymorphic type 1 HBGAs may occur in the course of RV evolution , explaining the complex P[II] genogroup that mainly causes diseases in humans but also in some animals .
Rotaviruses ( RVs ) are a major cause of severe gastroenteritis in children under the age of 5 , causing 200 , 000 deaths[1–3]and account for 2 million childhood hospital admissions with an estimated cost of over 1 billion US dollars per year[4 , 5] . It has been shown that RV attachment to cell surface carbohydrates , mediated by the VP8* domain of the spike protein VP4 , is a required first step of an effective infection [6–9] . RVs are genetically diverse . Based on the VP4 sequences , the group A RVs have been grouped into 40 genotypes ( P[1]-P[40] ) [10 , 11] . In a study based on the VP8* sequence , the different RV genotypes have been grouped into five genogroups ( P[I]-P[V] ) [12] . Different genotypes/genogroups cause diseases in different populations and/or various animal species and each genotype and genogroup may have distinct glycan binding specificities responsible for their host ranges or tropism . The previously identified “sialidase-sensitive” animal genotypes in the P[I] genogroup are the first examples exhibiting such genotype specificities that require interaction with terminal sialoglycans for cell entry [13–15] . However , most other genotypes in the five genogroups are found to be sialidase-insensitive and many of them have been found to recognize HBGAs [12 , 16–19] . Human HBGAs are highly polymorphic containing the ABO , secretor ( H ) and Lewis families with wide distributions in the world populations and therefore may significantly affect RV epidemiology and disease burden . For example , the P[III] RVs that infect humans and many animal species have been found to recognize the type A HBGAs that are shared between humans and many animal species [12] . The P[11] genotype in P[IV] has been found to recognize the type 2 HBGAs precursors , which may be responsible for their age-specific host ranges in neonates and young infants due to the stepwise synthesis of HBGAs that is developmentally regulated [18 , 19] . The P[11] RVs also commonly infect bovines which is believed to be due the occurrence of the specific HBGA precursors in these animals [20] . Significant advances of the molecular basis of ligand-controlled host ranges of RVs have been made following crystallographic studies of the RV VP8*s in complex with their genotype-specific ligands , including the sialic acid dependent animal RVs P[3] and P[7] , the A antigen binding P[14] that infect both humans and animals , and P[11] that recognizes the type 2 precursor and commonly infects neonates , young infants and some animals[16 , 20–22] . Interestingly , the VP8*s from all these animal and human RVs adopt a similar galectin-like fold , and interact with a specific glycan in the cleft region although the P[11] cleft is wider than in the other three RVs[20] . In addition , while the glycan binding site of P[11] VP8* has moved to span almost the entire length of the cleft , the P[14] and P[3]/P[7] RVs shared a common binding site located in one corner of the cleft region . These data are valuable for our understanding of RV host ranges , RV evolution and particularly zoonosis as many RVs infect both humans and animals . Despite recent progress in elucidating VP8*/receptor interactions , the molecular basis of host ranges of the major human RVs ( P[4] , P[6] and P[8] ) in the P[II] genogroup that are responsible for over 90% of human infections remains unknown due to lack of a conclusive data detailing the precise interactions of VP8* domains with host ligands as receptors for these human RVs[17 , 23 , 24] . Crystal structures of the native VP8* of the P[4] and P[8] RVs also showed a galectin-like fold with a similar wider cleft between the two β-shifts as that of P[11] RVs , which has led to a deduction of a similar cleft-glycan-binding site for P[4] and P[8] RVs[16 , 20 , 22 , 25] . However , in our recent study of another P[II] RV ( P[19] ) that is genetically closely related with the other P[II] RVs , we found that P[19] VP8*s have a unique property of binding two types of glycans , the mucin core and type 1 histo-blood group antigens ( HBGAs ) and may use a binding site different from those described above[26] , indicating that the P[II] RVs may be evolutionarily distantly related with other genogroups . To seek direct evidence on the possible shifted ligand binding site of P[19] VP8* and to explore the molecular basis how could a single binding site accommodate two structurally related but distinct glycan ligands , in this current study we resolved the crystal structures of the P[19] VP8* in complex with the two glycans . The structural data showed that the P[19] VP8* uses a completely new binding site that is different from the one of other RVs and this single binding site is able to interact with either type I HBGA pentasaccharide Lacto-N-fucopentaose I ( LNFP I ) or mucin core 2 glycan with a similar binding mode . Structure based-sequence comparisons also confirmed the conserved binding site of P[19] with other P[II] RVs . Unlike the other three P[II] RVs that mainly infect humans , the P[19] RVs are rarely found in humans but commonly infect animals ( porcine ) . Thus , our study helps establish the genetic and evolutionary relationships among these human and animal RVs , which advanced our understanding of RV host ranges , disease burden , epidemiology , and zoonosis of human diseases .
Previous glycan array studies have shown that the P[19] VP8* recognizes the mucin core glycans with a key GlcNAcβ1-6GalNAc motif and the type 1 HBGA precursor with inclusion of the internal Gal ( Fucα1-2Galβ1-3GlcNAcβ1-3Gal ) [26] . Prior to crystallographic studies , the binding specificity of P[19] VP8*was validated by ELISA using recombinant GST-tagged VP8* with the type 1 HBGA penta-saccharide LNFP I and mucin core 2 glycan ( Fig 1 ) . The binding of P[14] VP8* to A-type HBGAs was included as a positive control . Our results showed that the binding signals of P[19] VP8*exhibited typical dose-responses to both LNFP I and mucin core 2 glycans . As expected in the controls , the VP8* of a P[14] RV ( P[III] ) only recognized the A antigen but neither the LNFP I nor mucin core 2 . The crystal structures of P[19] VP8* in complex with LNFP I and mucin core 2 were solved at 1 . 94 Å and 1 . 90 Å , respectively . The electron density was clear for both bound ligands , which allowed unambiguous assignment of the two ligands ( S1 Fig ) . The secondary structural elements from N- to C- termini are designated as: βA ( 73–74 ) , βB ( 80–85 ) , βC ( 90–96 ) , βD ( 102–108 ) , βE ( 115–121 ) , βF ( 124–130 ) , βG ( 137–144 ) , βH ( 152–159 ) , βI ( 163–169 ) , βJ ( 172–177 ) , βK ( 184–189 ) , βL ( 197–200 ) , βM ( 204–208 ) , and αA ( 212–221 ) ( S2 Fig ) . In comparison with the previously solved P[19] VP8* native structure [27] , ligand binding did not cause any significant conformational changes , with the root mean squared deviation ( RMSD ) for alpha carbons of the backbone atoms between the bound P[19] VP8* and the free VP8* being 0 . 531 Å ( mucin core 2 ) and 0 . 505 Å ( LNFP I ) , respectively . Interestingly , a structural rearrangement was noted among residues 87–90 in the B-C loop after P[19] VP8* bound to either of the glycans ( S3 Fig ) , with the RMSD for alpha carbons of these residues between bound and free P[19] VP8* being 1 . 070 Å ( mucin core 2 ) and 0 . 992 Å ( LNFP I ) , respectively . Similar to other known VP8* structures , P[19]VP8*s adopted a classical galectin-like fold with two twisted antiparallel β-sheets consisting of strands A , L , C , D , G , H and M , B , I , J , K , respectively ( S2 Fig ) . The common shallow cleft between the two β-sheets where the glycan-binding sites of other known RVs are located ( Fig 2a–2d ) is wide in P[19] ( Fig 2a and 2e ) similar to the P[11] and P[8]/P[4] VP8*s but wider than the P[3]/P[7] and P[14] VP8*s . It was noted that the P[19] glycan binding site is located away from the cleft and thus represents a completely new glycan binding site , consistent with our previous observations based on NMR and mutagenesis studies [26] . This new glycan-binding site is composed of the carboxyl-terminal α-helix and the β sheet that composed of the βB , βI , βJ , and βK strands ( Figs 2e and 3 ) . The residues involved in the LNFP I interaction include W81 , L167 , H169 , G170 , G171 , R172 , W174 , T184 , T185 , R209 , E212 , and T216 , while those involved in mucin core 2 interactions are W81 , L167 , H169 , G170 , G171 , R172 , W174 , T185 , R209 , and E212 ( Fig 4 ) . LNFP I binds P[19] VP8* with a network of hydrogen-bonding interactions and hydrophobic interactions ( Fig 3 ) . All four residues in the type 1 HBGA chain backbone , Galβ1-3GlcNAcβ1-3Galβ1-4Glc participated in the interaction , with the motif Galβ1-3GlcNAcβ1-3Gal playing a central role . The fucose in the penta-saccharide LNFP I , referred as Fuc-I , lies almost 90 degree relative to the plane formed by the rest of LNFP I residues and thus projects out from the VP8* binding surface without making direct contacts with VP8* . The galactose next to Fuc-I , referred as Gal-II , interacts with residues T184 and T216 via hydrophobic interactions , which is further stabilized by forming hydrogen-bonds with T185 and two water molecules . The GlcNAc at the third position ( GlcNAc-III ) inserts into the deep binding pocket formed by W81 , L167 , W174 , T185 , R209 , and E212 , forming hydrophobic interactions between the LNFP I backbone and residues W81 , L167 , W174 , and E212 , and forming hydrogen bonds between the LNFP I and the side chains of residues T185 , R209 , and E212 ( Fig 3b ) . Two other water molecules help stabilize GlcNAc-III through hydrogen bonds with its acetyl moiety . The fourth saccharide ( Gal-IV ) makes contacts with G170 , G171 , and W174 through hydrophobic interactions and forms hydrogen bonds with H169 , G170 and two water molecules . The fifth saccharide ( Glc-5 ) interacts with R172 through hydrophobic effects . P[19] VP8* binds mucin core 2 using the same binding site ( Fig 4 ) through an almost identical conformation as that of bind LNFP I , with a RMSD 0 . 118 Å for the alpha carbon backbone atoms . Two amino acids , T184 and T216 , which are involved in binding to LNFP I , did not participate in binding to the shorter mucin core 2 trisaccharide ( Fig 5 ) . The GalNAc ( GalNAc-I ) forms hydrogen bonds with G171 and R172 , while the Gal ( Gal-II ) makes hydrophobic interactions with G170 and G171 . The amino acid residue of threonine that links to GalNAc-I pointed into the solvent . Interestingly , the GlcNAc ( GlcNAc-III ) inserted into the same deep binding pocket that binds the GlcNAc-III of LNFP I ( Fig 3 ) . It was noted that H169 contributed to the mucin core 2 binding interaction with GlcNAc-III , but this interaction was not seen with GlcNAc of LNFP I ( Fig 5 ) . The GlcNAc-III was further stabilized by two amino acids ( T185 , R209 ) through hydrogen bonding interactions . The superimposition of LNFP I and mucin core 2 in interaction with P[19] VP8* and the schematic interacting diagram was shown in Fig 5 . To examine the biological significance of reactive glycan ligands to P[19] RV function , viral replication inhibition assays were performed on a human P[19] RV . P[19] RV titers were significantly reduced following incubation of the viruses with combination of both LNFP I and mucin core 2 ( Fig 6 ) . Structural-based sequence alignment of P[4] , P[6] , P[8] and P[19] among P[II] RVs showed significant amino acid conservation on the newly identified ligand binding interface ( Fig 7 ) . Residues W81 , L167 , W174 , T184 , T185 , R209 , and E212 are identical among all four P[II] genotypes based on representative sequences from each genotype , while there are slight difference at residues 169 , 170 , 171 , 172 , and 216 . For example , the H169 in P[6] and P[19] is changed to Y169 in P[4] and P[8] , and the R172 in P[4] , P[8] and P[19] is changed to S172 in P[6] . As LNFP I and mucin core 2 share the same binding interface in P[19] VP8* , we use LNFP I as a representative to investigate whether P[4] and P[8] could accommodate the ligand in the newly found glycan-binding site as their native crystal structures are available . The overall structures of the LNFP I-bound P[19] VP8* domain and the apo P[4] VP8* domain ( PDB: 2AEN ) were very similar based on the backbone alpha carbon RMSD of 0 . 569 Å ( Fig 8a ) . Superimposition of the P[4] VP8* apo structure onto the P[19] VP8* structure in its complex with LNFP I showed that the LNFP I fit almost perfectly onto the surface of the apo P[4] VP8* defined by the same residues observed in the binding interface of LNFP 1 to the P[19] VP8* ( Fig 8a , inset ) . The only notable differences were that the sidechains of Y169 and R209 in the P[4] VP8* tilted away from the LNFP 1 binding pocket , which may destabilize the interaction between LNFP I and P[4] VP8* . The LNFP I-bound P[19] VP8* also had a similar overall structure as the apo P[8] VP8* domain ( PDB: 2DWR ) , with the RMSD of the alpha carbons of the backbone being equal to 0 . 493 Å ( Fig 8b ) . Again , superimposition of the two structures indicated that the side chains of almost all of the residues involved in binding to LNFP I in the P[19] VP8* domain were also in position to interact with the LNFP 1 except that the side chain of R172 in the P[8] VP8* domain sterically clashed with the oxygen atom of the Gal-IV moiety of LNFP I in the superimposed structure , which likely is responsible for preventing P[8] VP8* from binding to LNFP I . These structural analyses suggested that P[4] and P[8] as well as P[6] ( the crystal structure of P[6] remains unavailable ) may all utilize the same ligand binding site identified in P[19] , however , slight sequence and structural variations among the VP8* domains encoded within the different genotypes may be responsible for genotype-specific ligand-binding patterns and specificities of different genotypes .
Our crystallography studies clearly demonstrated that the P[19] RVs use a novel glycan binding site that is different compared to known binding sites of other RV genotypes and genogroups . This new binding site is capable of interacting with two structurally related but distinct glycans , the mucin core 2 and type 1 HBGA of LNFP I , using a common binding pocket and similar binding mode . Sequence comparisons showed high levels of amino acid conservation of the P[19] binding site with other P[II] genotypes ( P[4] , P[6] and P[8] ) and two genotypes in P[I] . These data confirmed the new binding site of P[19] previously deduced from NMR and mutagenesis analyses[26] , supporting the hypothesis that P[II] RVs are under strong selection of the host mucin core and/or type 1 HBGAs as common traits . However , given the complicated P[II] genogroup with multiple genotypes and variable host ranges among different human populations and/or some animal species , further defining the molecular basis of how these two structurally related host ligands drive RV evolution leading to such diverse P[II] genogroup is important . Both mucin cores and HBGAs are O-linked glycans commonly seen in nature on the mucosal surfaces and cellular membranes of many mammals . These O-linked glycans are synthesized step-wisely by a group of glycosyltransferases , in which the GlcNAc-containing oligosaccharide motifs that are recognized by P[II] RVs serve as the starters or precursors ( S5 Fig ) . For example , the type 1 HBGA precursor Galβ1-3GlcNAc can be extended to different A , B , H and Lewis HBGA products by adding one specific saccharide in each step . This process is developmentally regulated in the early lives of children [18 , 20 , 26] , which may also occur in animals , leading to shared precursors , intermediates and/or full HBGA products in some animal species . Given the high sequence conservation constituting the GlcNAc-binding pocket among all P[II] RVs , we deduced that the P[II] genogroup may originate from an ancestor with a simple binding site recognizing these GlcNAc-containing motifs ( GlcNAc-Gal ) and circulating in one or a group of species that share such glycans . Such a binding site could further expand its receptor binding repertoire through genetic variations in the course of RV evolution by adapting to additional residues when encountering new hosts producing longer and more complicated HBGA products or along the course of RV-host co-evolution . The above deduction assumes that , while the binding ability to the GlcNAc-containing moiety is maintained , extended interactions with additional saccharides could affect receptor ligand binding affinity and/or specificity , therefore potentially changing the binding outcomes . This assumption is supported by the observed variable binding patterns of the four P[II] RVs ( P[4] , P[6] , P[8] , and P[19] ) to the tetra- ( LNT ) , penta- ( LNFP I ) and hexa-saccharide ( LNDFH I ) of the type 1 HBGAs that either contain ( LNDFH I ) or do not contain ( LNT and LNFP I ) the Lewis epitopes[26] . The observed inability of P[4] and P[8] binding to the penta-saccharide LNFP I without the Lewis epitope is further supported by homology modeling of the native P[4] and P[8] VP8* structures in comparison with the LNFP I bound P[19] VP8* with an identification of binding clash for both P[4] and P[8] VP8* ( Fig 8 ) . In addition , the amino acids H169 and T216 of P[19] that are involved in interactions with the residue Gals next to GlcNAc have changed to Y169 and N216 , which may also lead to different binding outcomes in the two genotypes . Finally , it was noticed that there is a slightly shifted orientation of the GlcNAc-containing motif inside the binding pocket between mucin core 2 and the type 1 HBGAs , leading to a significant orientation shift of the backbones between mucin core 2 and type 1 HBGA within the binding cleft ( Fig 5a ) . This indicated a mechanism for how P[19] VP8* is able to achieve a broader binding specificity to the extended molecules of the two glycan types . While the majority of P[I] RVs infect animals , most of the P[II] RVs infect humans . Thus , we deduced that the P[II] RV may come from P[I] with an animal host origin and were introduced to humans by adapting to the polymorphic human HBGAs , leading to different P[II] genotypes infecting different human populations and/or some animal species depending on their evolutionary stages . For example , the P[19] genotype may represent an early evolutionary stage after the ancestors of P[II] genogroup started adapting to human receptors but they may still retain the binding specificities to the backbones of the mucin core 2 and type 1 HBGAs . This may explain why the P[19] RVs are commonly found in animals ( porcine ) but rarely in humans . On the other hand , the P[4] and P[8] RVs are genetically closely related and both genotypes are well developed that recognize the much more matured HBGA product , the Lewis b ( Leb ) antigen that is widely distributed in humans , and together these two genotypes are responsible for ~90% of human RV infections worldwide . Furthermore , the P[6] RVs have been found to recognize the much less matured type 1 HBGA precursors , which is consistent with the fact that the P[6] RVs commonly infect neonates and young infants through recognizing the age-specific precursor glycans that commonly occur in the early lives of children [28 , 29] . The P[6] RVs are also commonly found to infect porcine , likely through the type 1 HBGA precursor glycans that are share between humans and pigs . Thus , the elucidation of such genotype-specific host ranges controlled by the host HBGA makeups is important for understanding the disease burden and epidemiology therefore vaccine strategy against RVs based on the P type vaccine approach [30] . The findings of sequence conservation of P[10]/P[12] binding sites with that of the P[19] and similar glycan binding profiles between P[10] and P[19] [26] extends our understanding of P[II] evolution , in which these two P[I] RVs may represent even earlier ancestors than P[19] of the P[II] lineage . Both P[10] and P[12] are minor genotypes occasionally found in humans and bats and horses , respectively [31 , 32] , while the majority of other P[I] RVs were more commonly found to cause diseases in different animal species , indicating that the P[10]/P[12] RVs are unique in P[I] and should be evolutionarily grouped to the P[II] lineage . In fact , P[10]/P[12] were genetically closer to P[19] , P[8]/P[4] and P[6] than the rest genotypes based on the full VP4 sequence phylogeny analyses [33 , 34] . Thus , the P[10]/P[12] RVs are considered to be the earliest traceable ancestor of P[II] lineage . The reason of low abundance of these two genotypes in any species remains unknown . In conclusion , P[II] RVs represent a unique evolutionary lineage starting from an ancestor in P[I] with a possible animal host origin . While the original binding specify to the mucin core and type 1 HBGA precursors is maintained , additional interactions with adjacent residues may have occurred during the evolution of the ancestor genotype as it adapted to human receptors . This led to the diversity of RV strains seen today , with some mainly infecting animals with others mainly infecting humans [26] . Since viruses in all group A RV genotypes and genogroups must be from a single ultimate common ancestor , the deduced evolutionary path of animal-to-human transition from P[I] to P[II] , but not the other way around , may apply to other genotypes and genogroups . This deduction is important for RV classification and epidemiology , which may impact prevention and control strategies , such as vaccine design against RVs . For example , since the majority of the genotypes in P[I] exclusively infect animals , they may not be suitable for developing live human vaccine , because they may not be able to replicate in human guts due to the lack of proper receptor . This issue is urgent as the Jennerian approach is still widely used to develop live animal reassortant RV vaccines in many countries . Our research still has certain limitations . For example , the deduced binding sites for the major human RVs P[4] , P[6] and P[8] remain to be verified by co-crystallization studies of the VP8* in complex with their ligands . This task is challenging as our homology model data indicated that the free VP8* does not accept Lewis b , consistent with previous observations from several other groups [23 , 24] . In addition , the precise saccharide sequences and structures recognized by the major P[II] RVs remain unknown . As the cleft where the P[19] binding site is located is long and fairly deep , it is likely to accommodate a long glycan extending to both sides of the GlcNAc-containing oligosaccharide motifs , and future studies to explore this issue by glycan arrays with more representative human glycan pools are necessary . Finally , it is noted that the new binding site of P[19] shifts toward the C terminus of VP8* and is located in the bottom of VP8* , which may need the support of VP5* to exhibit its true binding characteristics . This issue also needs to be studied .
The VP8* core fragments ( amino acids 64 to 223 ) of the human RV P[19] with an N-terminal glutathione S-transferase ( GST ) tag was overexpressed in Escherichia coli BL21 ( DE3 ) cells as previously described [12 , 17] . Cells were grown in 1L Luria broth ( LB ) medium supplemented with 100 μg ml-1 ampicillin at 310 K . When the OD600 reached 0 . 8 , 0 . 5 mM isopropyl-β-D-thiogalactopyranoside was added to the medium to induce protein expression . The cell pellet was harvested within 12 h after induction and re-suspended in the 30 ml phosphate-buffered saline ( PBS ) buffer ( 140 mM NaCl , 2 . 7 mM KCl , 10 mM Na2HPO4 , 1 . 8 mM KH2PO4 , pH 7 . 3 ) . The cells were lysed by French press ( Thermo Fisher Scientific , Waltham , MA ) , then the cell debris was removed by centrifugation at 12 , 000×g for 30 min . The supernatant of the bacterial lysate was loaded to a disposable column ( Qiagen , Hilden , German ) pre-packed with glutathione agarose ( Thermo Fisher Scientific ) . After three washes with PBS buffer , the GST fusion protein of interest was eluted with elution buffer ( 10 mM reduced glutathione , 50 mM Tris-HCl , pH 8 . 0 ) . The GST tag of the VP8* protein was removed using the thrombin ( Thermo Fisher Scientific ) after dialysis into the buffer ( 20 mM Tris-HCl , 50 mM NaCl , pH 8 . 0 ) . The flow-through was collected after passing the mixture through glutathione agarose , and the purified protein was concentrated to about 10 mg ml-1 with an Amicon Ultra-10 ( Millipore , Billerica , MA ) . The purity of the protein was judged using 15% SDS-PAGE stained with Coomassie Brilliant Blue . The binding of purified VP8* to different glycans was confirmed by ELISA . Synthetic polyacrylamide polymer ( PAA ) conjugated oligomers were used to study their specificity as a ligand for P[19] RVs . Briefly , microtiter plates ( Thermo Fisher Scientific ) were coated with recombinant VP8* proteins ( 10 μg/ml ) at 4°C overnight . After blocking with 5% nonfat cow milk , synthetic polyacrylamide polymer ( PAA ) -biotin conjugated oligomers ( Glycotech , Gaithersburg , MD ) were added at serial dilutions and incubated at 4°C overnight . The bound oligosaccharides were then detected using HRP-conjugated-streptavidin ( Jackson Immuno Research Laboratories , West Grove , PA ) and displayed using the TMB kit ( Kierkegaard and Perry Laboratory , Gaithersburg , MD ) . The hanging-drop vapor-diffusion method was used for crystallizing human RV P[19] VP8* protein and co-crystallizing VP8* complexed with LNFP I and mucin core 2 . Crystals were obtained from drops where 1 μl purified P[19] VP8* was mixed with 1 μl of the reservoir buffer: 0 . 5 M ammonium sulfate , 0 . 1 M sodium citrate tribasic dihydrate pH 5 . 6 , 1 . 0 M lithium sulfate monohydrate . The sugar LNFP I ( Dextra , Reading , UK ) and threonine linked mucin core 2 ( kindly provided by James C . Paulson at the Scripps Research Institute ) were prepared at 60 mM and 100 mM , respectively , in PBS with 10% glycerol and soaked into the crystals . Diffraction data were collected at the Advanced Photon Source ( APS ) beamline 31-ID-D , Argonne National Laboratory , in Chicago , Illinois . A total of 360 images were collected using 0 . 5° oscillation during 20 s exposures . The images were integrated with MOSFLM [35] , and scaled with SCALA [36] . Molecular replacement was performed with PHASER [37] using the coordinates of chain A from 5GJ6 [27] as the search model . Iterative model building was manually carried out in COOT [38] , and refinements using 5% of reflection in Free-R set were carried out in REFMAC [39] implemented in the CCP4 suite [40] . The structure quality was assessed using Mol Probity [41] . Final model and scaled reflection data were deposited at the Protein Data Bank ( PDB ID 5VKS and 5VKI ) . Processing and refinement statistics for the final model are presented in Table 1 . The visualization and investigation of the final model was analyzed using Chimera [42] . The P[19] RVs ( strain 210 ) which was cell-culture adapted by multiple blind passages on the MA104 cells ( passage 4–6 ) were then used for inhibition assays with the oligosaccharides LNFP I and mucin core 2 using procedures described previously[26] . The P[19] RVs at 300 fluorescent forming units ( FFU ) /10 μl ) were pre-incubated with different inhibition reagents for 30 min . After rinsing twice with serum-free DMEM and chilling of all reagents and the 24-well plates on ice , duplicated wells of confluent MA104 monolayers were inoculated with the virus-oligosaccharide on ice with continuous rocker platform agitation for 1 h . The inoculum was then removed and the cells were washed twice with ice-cold serum-free DMEM . The plates were then placed back in the 37°C incubator for 18 to 20 h prior to quantification of infected cells by immunofluorescence with a rabbit anti-rotavirus antibody followed by a FITC-labeled goat anti-rabbit secondary antibody . | Rotaviruses ( RVs ) are diverse , infecting humans and/or animals . Significant advances in understanding ligand-associated RV host ranges have been made but how such host ligands drive RV evolution leading to the diverse genotypes/genogroups already identified remains unclear . In this study , through solving crystal structures of P[19] VP8*-ligand complexes with two different ligands , we demonstrated how genetic variations could configure a totally new ligand binding site leading to a distinct new evolutionary lineage . Sequence comparison also identified further changes of the binding site which may occur over the course of RV evolution leading to different P[II] genotypes infecting different populations , including some animal species widely seen everywhere around the world . The elucidation of the genetic and evolutionary relationships among all members of the P[II] lineage including the two genotypes in P[I] is highly significant for advancing our understanding of RV host ranges , disease burden and zoonosis of human diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"organismal",
"evolution",
"chemical",
"bonding",
"medicine",
"and",
"health",
"sciences",
"crystal",
"structure",
"pathology",
"and",
"laboratory",
"medicine",
"condensed",
"matter",
"physics",
"microbiology",
"microbial",
"evolution",
"crystallography",
"hydrogen",
"bonding",
"physical",
"chemistry",
"infectious",
"diseases",
"solid",
"state",
"physics",
"proteins",
"chemistry",
"pathogenesis",
"rotavirus",
"infection",
"evolutionary",
"genetics",
"viral",
"evolution",
"physics",
"biochemistry",
"mucin",
"host-pathogen",
"interactions",
"virology",
"animal",
"evolution",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"viral",
"diseases",
"evolutionary",
"biology"
] | 2017 | Structural basis of glycan specificity of P[19] VP8*: Implications for rotavirus zoonosis and evolution |
The avian H7N9 influenza outbreak in 2013 resulted from an unprecedented incidence of influenza transmission to humans from infected poultry . The majority of human H7N9 isolates contained a hemagglutinin ( HA ) mutation ( Q226L ) that has previously been associated with a switch in receptor specificity from avian-type ( NeuAcα2-3Gal ) to human-type ( NeuAcα2-6Gal ) , as documented for the avian progenitors of the 1957 ( H2N2 ) and 1968 ( H3N2 ) human influenza pandemic viruses . While this raised concern that the H7N9 virus was adapting to humans , the mutation was not sufficient to switch the receptor specificity of H7N9 , and has not resulted in sustained transmission in humans . To determine if the H7 HA was capable of acquiring human-type receptor specificity , we conducted mutation analyses . Remarkably , three amino acid mutations conferred a switch in specificity for human-type receptors that resembled the specificity of the 2009 human H1 pandemic virus , and promoted binding to human trachea epithelial cells .
The 2013 avian H7N9 virus outbreak in China was tied to human exposure to infected poultry in live bird markets [1] . Closure of the markets halted new human infections , but upon reopening , several additional outbreaks occurred; 779 human infections have been documented to date according to the WHO [2] . While there are reports of possible human-to-human transmission [3–5] , H7N9 has not acquired the capability for sustained transmission in the human population . Receptor specificity of influenza A viruses is widely considered to be a barrier for transmission of avian influenza viruses in humans [6] . Over the past 50 years , the strains circulating in the human population include the H3N2 strain that caused the 1968 pandemic , a seasonal H1N1 strain introduced in 1977 , and an H1N1 pandemic strain that emerged in 2009 and replaced seasonal H1N1 viruses . All human pandemic strains to date have exhibited specificity for human-type receptors ( α2–6 linked ) , in contrast to their avian virus progenitors that recognize avian-type receptors ( α2–3 linked ) [7 , 8] . In each case , the change in receptor specificity from avian-type to human-type involved two mutations in the HA receptor binding pocket , E190D and G225D for the H1N1 viruses , and Q226L and G228S for the H2N2 and H3N2 viruses [9 , 10] . These insights have framed current efforts to determine how avian influenza with other HA serotypes might acquire human-type receptor specificity . For the H5N1 HA , introduction of the two H1 specificity-switching mutations abolished receptor binding altogether , while the H3 mutations retained avian-type receptor binding , with minimal effect on receptor specificity [11–15] . However , introducing the Q226L mutation in combination with other mutations both increased binding to human receptors and conferred respiratory droplet transmission in ferrets [16–18] . While the H7N9 virus with the Q226L mutation maintained receptor specificity for avian-type receptors , some increase in avidity for human-type receptor analogs was noted [19–21] . We therefore reasoned that additional mutations might enable a full switch to human-type receptor specificity [22] .
We undertook a systematic mutation analysis of conserved residues in the H7 receptor-binding pocket . In addition to assessing the residues that conferred a receptor switch in the H1 and H3 hemagglutinins ( Fig 1A ) , we focused on three other residues that might impact binding of human-type receptors . 1 ) In the crystal structure of H7 HA , we noted that the positively charged side chain of K193 points directly into the binding pocket [20] . This could potentially inhibit binding of extended α2–6 sialosides that are known to project over the face of the 190-loop [23] . This position is invariably a threonine or serine in human H2 and H3 viruses , respectively , and recently has been implicated to be important in the evolution of the H3N2 pandemic virus [24] . We also used molecular modeling to show that K193 would likely physically interfere with the portion of the receptor glycan that is projecting from the sialic acid bound to the receptor-binding domain ( Fig 1B ) . 2 ) In a study of tissue tropism of H5N1 ( A/Indonesia/05/05 ) , a V186K mutation was found to confer binding to human trachea tissue sections . The G186V mutation was also noted as a potential adaptation of avian H7 to human-type receptors [25 , 26] and , in the H2 HA , N186 has been documented to form a hydrogen bond network that enables human-type receptor binding [27] . 3 ) The N224K mutation was identified as a critical residue for aerosol transmission of an H5N1 virus [28] . Varied combinations of mutations were introduced into the A/Shanghai/2/2013 ( Sh2 ) gene and expressed as recombinant , soluble , trimeric HA proteins in HEK293S GnTI ( - ) cells [29] . Each recombinant HA was tested for relative avidity to α2–3 ( avian-type ) and α2–6 ( human-type ) sialoside polymers in a glycan microarray based ELISA-like assay ( Fig 2A ) [30 , 31] , and for receptor specificity using a custom glycan microarray comprising 135 sialosides with matched sets of linear receptor fragments , O-linked and N-linked glycans , each terminating in NeuAcα2-3Gal and NeuAcα2-6Gal sequences ( Fig 2B , for a complete list see S1 Table ) [32] . The wild-type Sh2 HA that contains the Q226L mutation has a high preference for avian-type receptors , with minimal binding to human-type receptors , as noted previously [20] . Introduction of the G228S mutation that is found in human H2 and H3 viruses retained binding to α2–3 sialosides , and gained significant but weaker binding to α2–6 sialoside in the ELISA-like assay ( Fig 2A ) . However , there was no binding to human-type receptors in the glycan array ( Fig 2B ) , which exhibits higher stringency [12 , 16] . In contrast , mutations that confer human-type receptor specificity for H1N1 strains , E190D and G225D , alone or in combination showed no binding to sialosides in the glycan array ( S3 Table ) . We then introduced K193T in the G228S background . This Sh2 mutant bound almost equally well to avian-type and human-type receptors in both assays . Introduction of V186K in the K193T-G228S background , resulted in binding to human-type receptors in the ELISA-like assay , with some residual avian-type receptor binding . On the glycan array , this V186K-K193T-G228S mutant only bound human-type receptors , and displayed strikingly high specificity for α2–6 linked sialic acid found on extended N-linked glycans with 3 to 5 LacNAc repeats . A similar binding profile was also observed for an otherwise identical mutant containing V186G . The binding profile of these triple mutants is practically identical to pandemic H1N1 Cal/04/09 ( Fig 2B , bottom ) [32] , which is known to transmit efficiently between humans . Since most human infections to date have resulted from exposure to infected chickens in live poultry markets , we next investigated the impact of the receptor switch on binding to human and chicken airway tissues . Sh2 bound exclusively to chicken and not human trachea ( Fig 2C ) . The G228S mutant showed very strong binding to the chicken respiratory tract and very weak yet observable binding to human trachea at the base of the cilia . For the Sh2 K193T-G228S double mutant , we observed binding to goblet cells in chicken trachea and to the base of the cilia in human trachea , consistent with this mutant having dual receptor specificity . The V186K-K193T-G228S mutant also showed dual receptor binding on chicken and human trachea sections . A triple mutant changing only V186K to V186G retained human-type receptor specificity , but exhibited reduced avidity and increased specificity for binding to human trachea epithelium , which correspond to properties similar to those of Cal/04/09 pdmH1N1 . On analysis of the V186N and N224K mutations , we found that V186N in just the G228S background led to specific binding to human-type receptors in both the ELISA-like assay with sialoside polymers ( Fig 3A ) and the glycan array ( Fig 3B ) . With the addition of N224K in the V186N-G228S background , we observed a significant increase in binding . Thus , there are multiple ways for H7N9 to obtain human-type receptor specificity . The N224K mutant does not confer specificity to human-type receptors in other Sh2 backgrounds ( S3 Table ) , but does increase binding in human-specific Sh2 mutants ( S1 Fig ) . We conclude that a lysine at position 224 does not significantly alter receptor specificity , but does enhance the strength of binding , likely through a positive avidity contribution [28] . The influence of the K193T mutation on human-type receptor binding was of particular interest because K193S was shown to be an essential mutation for the H3 Hong-Kong 1968 pandemic , and K193T for H10N8 to obtain binding to human-type receptors [24 , 37] . In an ideal cis conformation , the human-type receptor would bind and project from the sialic acid binding site towards the 190 helix and has the potential to interact with amino acids of the 190 helix that frame the top of the receptor binding site [23] . Moreover , we have recently shown that H3N2 viruses , as well as pandemic H1N1 , exhibit preference for branched N-linked glycans that feature elongated LacNAc repeats extending over the 190 helix . These receptors project the second branch over the top of HA such that the second sialic acid can reach the receptor binding site of a second protomer in the same trimer [32] . Using molecular modeling , we investigated the possibility that the K193T mutation in Sh2 H7N9 HA would impact simultaneous binding of human-type receptors on complex N-glycans , as shown in Fig 6 . Here the low-energy conformation of the extended glycan chain produces a steric clash with the K193 , forcing LacNAc moieties to adopt a conformation projecting out of the receptor-binding site , and away from the 190 helix . Such a clash likely disfavors the preferred binding mode , where the rest of the glycan arches over the top of the HA surface . As a result , bidendate binding involving the simultaneous coordination of another branch of the glycan to a second protomer in the HA trimer is not possible ( Fig 6A ) . In contrast , simulations show that T193 interacts with LacNAc , enabling it to come closer to the HA , facilitating a bidentate interaction where the glycan is able to extend over the top of the trimer , thus effectively increasing avidity ( Fig 6B ) . We also determined the crystal structure of the Sh2 V186K K193T G228S with and without avian- and human-type receptors ( S2A and S2B Fig ) . The structures were virtually identical compared to the previously determined crystal structure of the Sh2 H7 HA protein ( Protein Data Bank [PDB] code 4N5J [20] ) . Moreover , in co-crystals with monomeric human-type ( LSTc ) and avian-type ( LSTa ) receptor analogs , electron density is seen only for the sialic acid , consistent with low-affinity binding of the monovalent receptor to the receptor site ( S3 Fig ) and preference of the mutants for extended biantennary glycans that offer the potential for bidentate binding .
We demonstrate here that several alternative three-amino-acid mutations ( V186G/K-K193T-G228S or V186N-N224K-G228S ) can switch the receptor specificity of the H7N9 HA from avian- to human-type , a property required for transmission in humans and ferrets [38 , 39] . Of these mutations , only isolated examples of 186G and 193N have to date been reported in H7 avian isolates . The mutants show profound loss of binding to avian-type ( α2–3 linked ) receptors , and increased binding to human-type ( α2–6 linked ) receptors in both glycan microarrays and glycan ELISA-type avidity assays . The mutants exhibit preferential binding to a subset of human-type receptors with extended branched N-linked glycans that terminate with NeuAcα2-6Gal reported to be present in N-linked glycans in human and ferret airway tissues [23 , 40 , 41] . Notably , this specificity for a restricted subset of human-type receptors is shared with recent H3N2 viruses , and the 2009 H1N1 pandemic virus . We have also recently observed that different sets of mutations switch the H6N1 and H10N8 HAs to human-type receptor specificity and , in each case , confer specificity for a similar subset of human-type receptors [37] ( de Vries , Tzarum , Wilson & Paulson manuscript in revision ) . Thus , recognition of human-type receptors with extended glycan chains appears to be a common characteristic of human influenza virus HAs , and avian virus HA mutants that bind to human-type receptors . Ideally , it would be important to assess the impact of the switch in receptor specificity in the ferret model that displays human-type receptors in the airway epithelium and is used to assess the propensity for air droplet transmission of human viruses . However , the introduction of the mutations that switch receptor specificity into an actual H7N9 virus background would represent gain-of-function ( GoF ) experiments that are currently prohibited [42] . During the course of this study , no viruses were created , and no experiments assessing the potential for air droplet transmission were performed . We suggest that understanding mutations that can confer human-type receptor binding will benefit risk assessment in worldwide surveillance of H7N9 in poultry and humans .
IRB & IACUC & IBC approval obtained at the funded institution . The tissues used for this study were obtained from the tissue archive of the Veterinary Pathologic Diagnostic Center ( Department of Pathobiology , Faculty of Veterinary Medicine , Utrecht University , The Netherlands ) . This archive is composed of paraffin blocks with tissues maintained for diagnostic purposes; no permission of the Committee on the Ethics of Animal Experiment is required . Anonymized human tissues were obtained under Service Level Agreement from the University Medical Centre Utrecht , The Netherlands . Use of anonymous material for scientific purposes is part of the standard treatment contract with patients and therefore informed consent procedure was not required according to the institutional medical ethical review board . " Codon-optimized H1 and H7 encoding cDNAs ( Genscript , USA ) of A/Shanghai/2/13 , Cal/04/09 and A/KY/07 were cloned into the pCD5 expression as described previously [29] . The pCD5 expression vector is adapted so that that the HA-encoding cDNAs are cloned in frame with DNA sequences coding for a signal sequence , a GCN4 trimerization motif ( RMKQIEDKIEEIESKQKKIENEIARIKK ) , and Strep-tag II ( WSHPQFEK; IBA , Germany ) . The HA proteins were expressed in HEK293S GnTI ( - ) cells ( ATCC ) and purified from the cell culture supernatants as described previously [43] . pCD5 expression vectors were transfected into HEK293S GnTI ( - ) cells using polyethyleneimine I ( PEI ) . At 6 h post transfection , the transfection mixture was replaced by 293 SFM II expression medium ( Gibco ) , supplemented with sodium bicarbonate ( 3 . 7 g/liter ) , glucose ( 2 . 0 g/liter ) , Primatone RL-UF ( Kerry ) ( 3 . 0 g/liter ) , penicillin ( 100 units/ml ) , Streptomycin ( 100 μg/ml ) , glutaMAX ( Gibco ) , and 1 . 5% DMSO . Tissue culture supernatants were harvested 5–6 days post transfection . HA proteins were purified using Strep-Tactin sepharose beads according to the manufacturer’s instructions ( IBA , Germany ) Purified , soluble trimeric HA was pre-complexed with horseradish peroxidase ( HRP ) -linked anti-Strep-tag mouse antibody ( IBA ) and with Alexa488-linked anti-mouse IgG ( 4:2:1 molar ratio ) prior to incubation for 15 min on ice in 100 μl PBS-T , and incubated on the array surface in a humidified chamber for 90 minutes . Slides were subsequently washed by successive rinses with PBS-T , PBS , and deionized H2O . Washed arrays were dried by centrifugation and immediately scanned for FITC signal on a Perkin-Elmer ProScanArray Express confocal microarray scanner . Fluorescent signal intensity was measured using Imagene ( Biodiscovery ) and mean intensity minus mean background was calculated and graphed using MS Excel . For each glycan , the mean signal intensity was calculated from 6 replicate spots . The highest and lowest signals of the 6 replicates were removed and the remaining 4 replicates used to calculate the mean signal , standard deviation ( SD ) , and standard error measurement ( SEM ) . Bar graphs represent the averaged mean signal minus background for each glycan sample and error bars are the SEM value . A list of glycans on the microarray is included in S1 Table . Purified HA trimers were precomplexed with anti-HIS mouse IgG ( Invitrogen ) and HRP-conjugated goat anti-mouse IgG ( Pierce ) , then diluted in series to required assay concentrations ( 40–0 . 05 μg/mL final ) . Preparation of streptavidin-coated plates with biotinylated glycans , incubation and washing of pre-complexed HA dilutions was exactly as described previously [23 , 32] . Sections of formalin-fixed , paraffin-embedded , human trachea and chicken trachea were obtained from the University Medical Center and the Department of Veterinary Pathobiology , Faculty of Veterinary Medicine , at Utrecht University , respectively . Tissue sections were rehydrated in a series of alcohol from 100% , 96% and 70% , and lastly in distilled water . Endogenous peroxidase activity was blocked with 1% hydrogen peroxide for 30 min at room temperature . Tissue slides were boiled in citrate buffer pH 6 . 0 for 10 minutes at 900kW in a microwave for antigen retrieval and washed in PBS-T three times . Tissue was subsequently incubated with 3% BSA in PBS-T for overnight at 4°C . On the next day , the purified HAs were precomplexed with mouse anti-strep-tag- HRP antibodies ( IBA ) and goat anti-mouse IgG HRP antibodies ( Life Biosciences ) at a 4:2:1 ratio in PBS-T with 3% BSA and incubated on ice for 20 minutes . After draining the slide , the precomplexed HA was applied onto the tissue and incubated for 90 minutes at RT . Sections were then washed in PBS-T , incubated with 3-amino-9-ethyl-carbazole ( AEC; Sigma-Aldrich ) for 15 minutes , counterstained with hematoxylin , and mounted with Aquatex ( Merck ) . Images were taken using a charge-coupled device ( CCD ) camera and an Olympus BX41 microscope linked to CellB imaging software ( Soft Imaging Solutions GmbH , Münster , Germany ) . The ectodomain of the Sh2 H7 HA mutant ( V186K-K193T-G228S ) was expressed in a baculovirus system essentially as previously described [20] . Briefly , the cDNAs corresponding to residues 19–327 of HA1 and 1–174 of HA2 ( H3 numbering ) of HA from A/Shanghai/2/2013 ( H7N9 ) ( Global Initiative on Sharing All Influenza Data ( GISAID ) isolate ID: EPI_ISL_138738 ) were codon-optimized and synthesized for insect cell expression and inserted into a baculovirus transfer vector , pFastbacHT-A ( Invitrogen ) with an N-terminal gp67 signal peptide , C-terminal trimerization domain , His6 tag , and thrombin cleavage site incorporated to separate the HA ectodomain and the trimerization and His tags . The HA1 domain triple mutant ( G228S , V186K , K193T ) was made by site-directed mutagenesis . The purified recombinant HA Bacmids were used to transfect Sf9 insect cells for overexpression . HA protein was produced in infecting suspension cultures of Hi5 cells with recombinant baculovirus at an MOI of 5–10 and incubated at 28°C shaking at 110 RPM . After 72 hours , Hi5 cells were removed by centrifugation and supernatants containing secreted , soluble HA proteins were concentrated and buffer-exchanged into 1xPBS , pH 7 . 4 . The HAs were recovered from the cell supernatants by metal affinity chromatography using Ni-NTA resin , and were digested with thrombin to remove the trimerization domain and His6-tag . The cleaved HAs were further purified by size exclusion chromatography on a Hiload 16/90 Superdex 200 column ( GE healthcare , Pittsburgh , PA ) in 20 mM Tris pH 8 . 0 , 100 mM NaCl , and 0 . 02% ( v/v ) NaN3 . Structure models were generated from PDBID 4LN8 [44] . A trimeric “head region” was created from residues K46 to S260 from the HA1 ( receptor binding region ) and residues Q61 to S93 from HA2 ( from the top of membrane fusion stem region ) . One structure was kept as WT , while each of the three binding sites in a second structure were altered by point mutations; G228S , V186K and K193T . The mutant model structures were generated in UCSF Chimera [45] , by selecting rotamers from the Dunbrack library [46] . A sialylated biantennary TriLacNAc N-glycan was generated on Glycam-Web ( www . glycam . org/cb ) and modeled into both the WT and mutated structure via computational carbohydrate grafting [47] , using the Neu5Ac in PDBID 4LN8 as a template . The reported grafting algorithm [48 , 49] was adapted to rotate the glycosidic linkages within normal bounds [50] , while monitoring the distance between the binding motif on the other arm of the glycan and the second HA binding site . The linkages were adjusted in series , beginning from the NeuAcα2-6Gal motif . A single optimal structure was selected based on the relative orientation and proximity of the NeuAcα2-6Gal motif on the other arm to the target HA binding site . The results were independent of whether the 3-arm or the 6-arm of the glycan was grafted onto the bound NeuAcα2-6Gal motif . The resulting structures were then subject to energy minimization and molecular dynamics simulation as described previously , to attempt to see whether the NeuAcα2-6Gal motif on the second arm of the glycan could locate into second binding site . Crystallization experiments were set up using the sitting drop vapor diffusion method . Initial crystallization conditions for the H7 mutant HA ( V186K-K193T-G228S ) were obtained from robotic crystallization trials using the automated CrystalMation system ( Rigaku ) at The Scripps Research Institute . Following optimization , diffraction quality crystals of the triple mutant HA were grown at 22°C by mixing 0 . 5 μl of protein ( 7 . 4 mg/ml ) in 20 mM Tris , pH 8 . 0 , 100 mM NaCl with 0 . 5 μl of a reservoir solution containing 0 . 2 M tri-potassium citrate , 5% ( v/v ) ethylene glycol and 22% ( w/v ) PEG3350 . The crystals were flash-cooled in liquid nitrogen by adding 20% ( v/v ) ethylene glycol to the mother liquor as cryoprotectant . The triple mutant HA-ligand complexes were obtained by soaking HA crystals in the well solution that now contained glycan ligands . Final concentrations of ligands LSTa ( NeuAcα2-3Galβ1-3GlcNAcβ1-3Galβ1-4Glc ) and LSTc ( NeuAcα2-6Galβ1-4GlcNAcβ1-3Galβ1-4Glc ) were all 5 mM , and soaking times were 10 min . Diffraction data were collected on synchrotron radiation sources specified in the data statistics tables . HKL2000 ( HKL Research , Inc . ) was used to integrate and scale diffraction data . Initial phases were determined by molecular replacement using Phaser [51] with the wild-type HA structure ( PDB codes 4N5J ) as a model . One HA protomer is present per asymmetric unit . Refinement was carried out using the program Phenix [52] . Model rebuilding was performed manually using the graphics program Coot [53] . Final refinement statistics are summarized in S2 Table . Thermal denaturation was studied using a nano-DSC calorimeter ( TA instruments , Etten-Leur , The Netherlands ) . HA proteins were eluted from the streptavidin beads in PBS with 2 . 5mM desthiobiotin , and 100μg of protein was tested . After loading the sample into the cell , thermal denaturation was probed at a scan rate of 60°C/h . Buffer correction , normalization , and baseline subtraction procedures were applied before the data were analyzed using NanoAnalyze Software v . 3 . 3 . 0 ( TA Instruments ) . The data were fitted using a non-two-state model . Atomic coordinates and structure factors have been deposited in the Protein Data Bank ( PDB ) under accession codes 5VJK , 5VJL and 5VJM for Sh2 mutant HA ( V186K-K193T-G228S ) in apo form and in complex with LSTc or LSTa . | Influenza A virus of the H7N9 subtype continues to cross the species barrier from poultry to humans . This zoonotic ability is remarkable as the virus retains specificity to avian-type receptors . To effectively transmit between humans , the virus needs to acquire human-type receptor specificity . In this study , we show that recombinant H7 proteins need three amino acid mutations to change specificity to human-type receptors . Although we are not allowed to assess if these mutations would lead to efficient transmission in the ferret model , this knowledge will aid in surveillance . If these amino acid mutations are observed to arise during natural selection in humans , timely actions could be taken . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"h7n9",
"chemical",
"characterization",
"medicine",
"and",
"health",
"sciences",
"microbial",
"mutation",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"vertebrates",
"orthomyxoviruses",
"animals",
"viruses",
"respiratory",
"system",
"h5n1",
"rna",
"viruses",
"h1n1",
"trachea",
"research",
"and",
"analysis",
"methods",
"receptor",
"binding",
"assays",
"infectious",
"diseases",
"influenza",
"a",
"virus",
"zoonoses",
"birds",
"medical",
"microbiology",
"microbial",
"pathogens",
"gamefowl",
"binding",
"analysis",
"fowl",
"poultry",
"anatomy",
"influenza",
"viruses",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"chickens",
"amniotes",
"organisms"
] | 2017 | Three mutations switch H7N9 influenza to human-type receptor specificity |
Bacterial cell-cell communication is mediated by small signaling molecules known as autoinducers . Importantly , autoinducer-2 ( AI-2 ) is synthesized via the enzyme LuxS in over 80 species , some of which mediate their pathogenicity by recognizing and transducing this signal in a cell density dependent manner . AI-2 mediated phenotypes are not well understood however , as the means for signal transduction appears varied among species , while AI-2 synthesis processes appear conserved . Approaches to reveal the recognition pathways of AI-2 will shed light on pathogenicity as we believe recognition of the signal is likely as important , if not more , than the signal synthesis . LMNAST ( Local Modular Network Alignment Similarity Tool ) uses a local similarity search heuristic to study gene order , generating homology hits for the genomic arrangement of a query gene sequence . We develop and apply this tool for the E . coli lac and LuxS regulated ( Lsr ) systems . Lsr is of great interest as it mediates AI-2 uptake and processing . Both test searches generated results that were subsequently analyzed through a number of different lenses , each with its own level of granularity , from a binary phylogenetic representation down to trackback plots that preserve genomic organizational information . Through a survey of these results , we demonstrate the identification of orthologs , paralogs , hitchhiking genes , gene loss , gene rearrangement within an operon context , and also horizontal gene transfer ( HGT ) . We found a variety of operon structures that are consistent with our hypothesis that the signal can be perceived and transduced by homologous protein complexes , while their regulation may be key to defining subsequent phenotypic behavior .
Comparing prokaryotic whole genome sequences to identify operons is a mature area of research [1] , [2] , [3] , [4] . Orthologous operon identification can imply a secondary degree of relation between components , reaffirming Clusters of Orthologous Groups ( COG ) and other assignments of function as well as suggesting essentiality [5] . This conservation of components also speaks to the conservation of signaling capacity in orthologous modular signaling operon-based units . That is , we are interested in ascertaining the genetic modularity of signal transduction processing , in particular those that operate within known , putative regulons . Drawing partly on previous work investigating microsynteny and gene neighborhoods [3] , [6] , [7] , we developed a general similarity search approach , we call a Local Modular Network Alignment Similarity Tool ( LMNAST ) . LMNAST applies a BLAST-like heuristic to gene order and arrangement . Resultant search hits help capture the conservation and phylogenetic dispersion of a given query modular network . Using , as queries , contiguously abutting genes of prokaryotic modular signaling networks , LMNAST identifies and scores hits based on the minimum number of frank mutations in gene organization needed to arrive at a given putative system homolog when starting from the query . Here , homology refers to similarity in relative gene order and relative transcriptional direction , after nucleotide level threshold filtering of gene elements based on BLAST [8] E-value . For the purpose of evaluation , two small modular systems were used as test inputs: one was the E . coli lac system and the other was the LuxS regulated ( Lsr ) system . In some ways , the two systems are quite similar ( Fig . 1 ) . Both import and catabolize the small molecules that induce system expression . For the lac system , this small molecule is , of course , lactose . For the Lsr system , the small molecule is autoinducer-2 ( AI-2 ) . AI-2 is a signaling molecule common among at least eighty bacterial species [9] . As mediated through either the Lsr or LuxPQ systems , bacteria are believed to use AI-2 to guide population based phenotypes , a phenomenon termed quorum sensing [10] . LuxPQ is a histidine kinase two component system , the regulon of which is distinct from Lsr and is not considered further . Lsr is an interesting query because its distribution should help elucidate its putative , modular quorum sensing function [9] and because the known homologs differ in gene organization [10] , [11] , [12] .
Input consisted of an ordered list of gene elements ( for example , lacIZYA ) . For each gene element a BLAST result file was generated using tblastn to search the nr/nt database for hits with E-values less than 0 . 1 , narrowing the search space . Each BLAST hit was assigned a character corresponding to the gene element queried . BioPerl [13] was used to query Genbank databases and process data from retrieved files . Nucleotide records with sufficiently proximal characters were investigated further . The degree of similarity between a putative hit and its corresponding query was evaluated according to the number of deletions , insertions , and rearrangements required to generate the putative hit from the query . Intra-hit gene duplications were disallowed as a simplification . Consequently , deletion could be noted by character type inclusion . Insertions of uncharactered elements between gene homologs were scored according to an affine gap rule whereby a portion of the deduction was scaled to the insertion length . Rearrangements refer to altered relative order and relative gene direction . Changed relative direction was only considered when relative order was maintained . When this criterion was satisfied , relative order was evaluated in terms of adjacent homolog distance , disregarding insertions and deletions . For each such structural dissimilarity there was a standard deduction in score . Noncontiguous elements were dropped iteratively until a maximum score was reached for each putative hit . When more than one putative hit version elicited equal scores in the same round , the version of the hit with the most characters was favored . Putative hits with scores greater than zero were retained . For evaluation purposes and to find a suitable balance between false positives and coverage completeness , each test query was run under both weak and stringent conditions . Stringent criteria searches assumed accurate annotation . Contrarily , weak criteria did not require genes to lie within the annotated coding sequence . Moreover , characters annotated as “pseudo” or bounded outside “gene” annotation were accepted as homologous characters . Weak criteria searches also allowed multiple genes to co-exist within the same annotation . Additionally , as a concession to the possibility of longer range interactions between genes , reduced gap penalties were used in weak criteria searches . Results described herein were derived using a gap penalty of 1 and 2 with an extension penalty of 0 . 3 and 1 , for weak and stringent criteria searches respectively . Mean element homology ( meH ) is a normalized , ancillary measure of string similarity as evaluated by BLAST . Useful for contrasting BLAST results to LMNAST hits , meH was calculated by normalizing each gene homolog's bit score to the maximum bit score for the entire corresponding BLAST result given a background subtraction of the minimum bit score . These normalized bit scores were then averaged for all gene elements within an LMNAST hit . A score of one indicates exact likeness whereas zero indicates the least degree of similarity . Also , widening the query beyond the system of interest to include a nominal number of flanking genes , here termed “extended window searching , ” afforded additional contextualization of LMNAST hit results . Finally , in evaluating certain low homology hits , nonscoring synonyms were used . Nonscoring synonyms are elements with equivalent gene annotation but insufficient homology according to the initial E-value filter . This is somewhat analogous to replacement in blastp .
We began evaluation of LMNAST by searching for the well characterized E . coli lac operon . Specifically , the E . coli lac genes lacI ( BAE76127 ) , lacZ ( BAE76126 ) , lacY ( BAE76125 ) , and lacA ( BAE76124 ) ( spanning bp 360473 to 366734 of the Genbank nucleotide record AP009048 ) were used as a query . The stringent criteria search yielded fewer hits than the corresponding weak criteria search ( 189 vs . 236 ) . Of the hits derived from the stringent criteria search , complete and perfectly arranged lac systems were found in 26 unique E . coli strains and S . enterica arizonae serovar 62:z4 , z23 ( meH 0 . 8 ) , the only Salmonella enterica serovar represented among all lac system hits , in keeping with its significant divergence from other serovars [14] . A representation of E . coli hits in a phylogenetic context is available in Fig . 3a . The average meH ( 0<meH<1 ) for these complete systems was 0 . 98 . An extended window query with five additional genes on either side of the original search frame , revealed eight complete systems with a hitchhiking , proximal cytosine deaminase after losing all other proximal genes . Only one system with all four characters was entirely removed from the original query's proximal gene set , suggestive of negligible stability for the canonical system outside of a limited phylogenetic domain . An additional 28 hits were bereft one lac system character ( average meH 0 . 74 ) . In all but three of these cases that missing gene was lacA . Of these hits , ten had an additional frank structural change to a divergent expression pattern originating between lacI and lacZ characters ( e . g . in E . cloacae ) , likely increasing system sensitivity to lacI repression in these cases [15] . Surprisingly , in other instances , extended window searching revealed the only proximal structural change to be a missing lacA gene . This lacA degeneracy may be indicative of its relative functional unimportance compared to other lac system members [16] . Some of the patterns described above can be inferred from coincidence heat charts ( Fig . 4 ) . These matrices represent LMNAST results by the frequency of coincidence between gene characters within hits . The shade of an index represents the frequency of hits where the row gene coincides with the column gene , normalized against the total number of hits containing the row gene , which itself is denoted by ( # ) . For example , in Fig . 4 , the left-most matrix is a representation of a theoretical set of homolog fragments ( AB , BC , CD , ABC , BCD , and ABCD ) . This simple set was constructed to only reflect unbiased homologous recombination presumably resulting only in chromosomal rearrangements . In this set , B and C were extant in five inputs , while A and D were extant in three inputs . All three inputs containing A also contained B , two also contained C , and one also contained D . This is reflected in the shades of the grids in the top row . The middle matrix represents the coincidence distribution among LMNAST E . coli lac hits . As an additional example , matrix element ( 2 , 1 ) is a rust color representing the 139 hits with a lacI character of the 155 also containing lacZ . Finally , the right-most matrix is the difference between the left and middle matrices . This particular analysis suggests , for example , that lacI is relatively over-represented across all hits , and that nearly all other coincidences are under-represented; surprisingly , this includes coincidences involving the permease , lacY . Unlike lacA , lacY is believed necessary for lactose catabolism , possibly pointing to the use of a lower affinity transporter in such cases . On the other hand , the over representation of lacI indicates an expected preference for the regulation of lactose catabolism . Of the strong criteria search results , 138 hits contained only two lac gene homologs ( average meH 0 . 28 ) . Two gene homologs represent the natural minimum of individual characters that a homologous system may contain . Such hits represented truncated systems , repurposed individual members , homoplasic convergence , or outright false positives . The majority of these hits fell within clusters of shared Genbank annotation in 2D similarity plots , which compare meHs ( averaged BLAST homologies ) against LMNAST homologies , or put differently , average amino acid identities against the system's broader organizational identity . Generically then , purely vertical displacements imply perfect conservation across species through either vertical or , more likely , recent horizontal gene transfer accompanied by amelioration , while purely horizontal displacements indicate recent gene loss and/or rearrangement . For purposes of downstream analysis , it is interesting to speculate that the kinetics of the remaining genes are unaffected in cases of purely horizontal displacement . For systems subject to HGT , such liberties must necessarily be taken with less confidence . In the case of the stringent lac search , similarity plots revealed a great deal of structural variability in the lac operon homologs of E . coli and near E . coli species ( Fig . 5 ) . Nonetheless , the canonical lac operon ( 26 ) and the paralogous evolved beta-galactosidase system ( 43 ) [17] are clearly the most dominant lac operon-homologs , perhaps partially reflecting the relative preponderance of fully sequenced E . coli strains . Addressing the full breadth of two character homologs , 87 contained lacZ and lacY character types , all of which were adjacent , five of which were misdirected relative to one another . Numerous truncated systems had high meH but imperfect organizational similarity . This cohort was restricted to strains of E . coli and closely related Shigella , Citrobacter , and Enterobacter species , reflecting a generally confined phylogenetic breadth among LMNAST lac hits ( Fig . 3b ) , and reinforcing the idea of limited lac horizontal gene transfer ( HGT ) [18] . The remainder of the hits consisted of adjacent repurposed characters with functional valence around sugar metabolism . This survey showed that LMNAST E . coli lac operon searches identified numerous ortholog and paralog instances . Relative disparities in gene preservation , gene loss , and structural rearrangements bearing signaling implications were delineated . While there was a significant degree of conformity to the standard genomic arrangement , the amount of diversity indicates that attention paid to related , non-canonical signaling units may be worthwhile . Further testing of LMNAST was conducted with weak , stringent , and expanded window searches of the E . coli Lsr system . The query Lsr system consists of a kinase ( LsrK: BAA15191 ) , a repressor ( LsrR: BAA15192 ) , ABC transporter genes ( LsrA: BAA15200 , LsrC: BAA15201 , LsrD: BAA15202 , and LsrB: BAE76456 ) , and phospho-AI-2 ( AI-2-P ) processing genes ( LsrF: BAE76457 , LsrG: BAE76458 ) . Along with AI-2 , the Lsr system consists of multiple overlapping positive and negative feedback loops . Multimeric LsrR represses system expression emanating from the intergenic region . AI-2-P , itself catabolized by LsrF and LsrG , allosterically relieves that repression . Thus , both expression troughs and peaks are tightly regulated [19] . For the LMNAST search we used the Lsr genes spanning bp 1600331 to 1609003 of E . coli K12 substrain W3110 ( Genbank nucleotide record AP009048 ) . The number of hits returned using stringent criteria totaled 419 . Much like the lac operon , the Lsr system appeared subject to imperfect conservation . Certainly , many fully sequenced E . coli bore exact Lsr homologs ( meH>0 . 95 ) . Exceptions were the truncated systems found in strains BL21 [20] , REL606 [20] , and E24377A , and the specific and complete excision of Lsr systems from an otherwise preserved gene order in B2 type E . coli ( Figs . 6 and S1 ) as revealed through expanded window searching . Unlike the lac operon , numerous Lsr system homologs had perfect LMNAST homology but markedly reduced meH ( Fig . 7a ) . This is suggestive of amelioration following recent HGT events ( which may itself be a reflection of a carefully tuned signal requiring the full complement and correct arrangement of Lsr elements ) . Indeed , Lsr system GC content varied in accordance with the background GC content , ranging from 0 . 35 to 0 . 71 . Finer scale GC analysis revealed a single consistent and curious feature across all hits with meH greater than 0 . 3: a sharply spiking dip in fractional GC content near the intergenic region ( Fig . S2 ) . This dip is suggestive of a conserved DNA binding domain essential to the signal transduction process , which would also , however , be a regulatory feature outside the scope of LMNAST searches . Imperfect LMNAST hits with meH greater than 0 . 3 , deviated from the theoretical distribution according to a bias towards the conservation of lsrB , F , and G , relative to the lsrA , C , and D importer genes ( Fig . 8 ) . This may be attributable to the fact that lsrB , F , and G likely pass cell signaling information downstream [14] , [21] , [22] , whereas loss of Lsr importer function might be partially redundant to a low affinity rbs pathway [23] , the likely alternate AI-2 import pathway [19] , [24] , [25] . In contrast to high meH systems , many systems with low meH ( <0 . 3 ) were involved in the metabolism of 5 carbon sugars , mainly ribitol and xylose , according to Genbank annotation ( Fig . 7b ) . Since AI-2 itself is mainly comprised of a 5 carbon ring , such homology is simultaneously intriguing and unsurprising . More generally among these low similarity hits , lsrK characters were commonly coincident with Lsr importer characters ( lsrA , C , D , and B ) , indicative of the functional link between such characters . These various features were laid more strongly in relief when measured against the proximal genetic background in an extended window search . While a representation of hit variability preserving structural information can be had from trackback plots ( Fig . S1 ) , additional salient results from stringent Lsr extended window searching could be deduced from the more summary coincidence heat maps ( Fig . 9 ) . The matrices indicate that lsrK and lsrA genes were strongly preserved among extended window hits . Also , if either lsrF or lsrG were present , the remaining Lsr genes were likely present . The complete system rescission mentioned before was hinted at , especially in rows 3 and 4 , corresponding to the toxin/antitoxin hipAB system . Intra-species variation of structural homology increased greatly when using stringent rather than weak criteria ( data not shown ) , mainly as a result of gene loss to pseudo gene conversion , mostly among transporter genes—a bias most easily explained as a matter of pure probability since there are more transporter genes than any other type , and a fact whose functional significance is blunted by the alternative AI-2 import pathway . These initial E . coli searches motivated other orthologous Lsr system queries . Full results for E . coli , S . Typhimurium , and B . cereus searches are available in Fig . S3 . These additional searches helped identify other possible Lsr system homologs , HGT partners , and non-canonical system-associated gene candidates . In Fig . S3 , we delineate operon directionality and gene homology . It is interesting to note that system variants exist among noted human pathogens: Yersinia pestis , Bacillus anthracis , and Haemophilus influenzae . In some instances , lsrRK are either absent ( e . g . E . coli BL21 ) or are associated with altered intergenic regions implying altered regulatory control ( e . g . Yersinia pestis Antiqua ) . In other cases transporter genes are distributed with altered bias due to position in the bidirectional operons ( e . g . Yersinia pseudotuberculosis PB1/+ ) . In some cases there is no LsrFG component ( e . g . Shigella flexerni 2002017 ) . LsrF and LsrG are AI-2-P processing enzymes that lower the intracellular AI-2-P level , thereby contributing to the repression of AI-2 induced genes . Given even only this modest degree of dispersion , it is nonetheless reasonable to suggest that the Lsr auto-induction system is , in fact , extant among scores of bacterial species and that because the organization of genes within the regulatory architecture is varied , the downstream phenotypic behaviors aligned with AI-2 regulated QS genes is likewise variable . Thus , our results are in line with a general hypothesis that the AI-2 quorum sensing system is broadly distributed and that the specific needs of the bacteria in a given niche are met by disparate operon arrangements . The overall phylogenetic distribution of the Lsr system mirrors that as developed by Pereira , et al . in the cluster they denote as Group I [26] . Here , however , details were fleshed out with different emphases . The Lsr LMNAST search captured the diversity of pseudo gene conversion , structural rearrangement , and additional hitchhiking genes associated with the Lsr systems that exist in the present nr/nt database . Moreover , inferences regarding regulatory Lsr system signals could be made that might also map to phylogenies or possibly , with much more effort , related ecological niches . Results from the various LMNAST searches were reconciled by taking the highest scoring hit among overlaps within each nucleotide record . In Fig . 10 , we overlaid LMNAST search results onto a phylogenetic tree [27] based on the E . coli genome and 16s data . Interestingly , Lsr system homologs clustered mainly in gammaproteobacteria with the greatest density being among E . coli strains . Diffusely manifesting in more distantly related bacterial species , the Lsr system appears to have been subject to several HGT events . That is , the Lsr system is absent in numerous Enterobacteriaceae species , while HGT gain events happened at the root of the Bacillus cereus group , to R . sphaeroides and R . capsulatus separately , to Sinorhizobium meliloti , and to Spirochaeta smaragdinae ( Fig . 10 ) . Curiously , while these bacteria occupy distinct ecological niches , they are all common to soil or water environments . Multiple extended window searches indicated that S . enterica was the most proximal cluster for every Lsr system HGT candidate . The sharing of a novel Lsr system-associated “mannose-6-phosphate isomerase” ( NP_460428 ) between Bacillus cereus group members , S . smaragdine , and S . enterica , further strengthened the suggestion of HGT partnership . The gene annotated as “mannose-6-phosphate isomerase” or “sugar phosphate isomerase , ” has recently been shown to be part of the LsrR regulon in Salmonella [28] . Although not part of the E . coli regulon , it was also associated with S . smaragdinae and B . cereus group orthologs . In keeping with a possible AI-2-P processing role , it was consistently adjacent to lsrK . Among gammaproteobacteria , parsimony suggests that two gain events of the Lsr system occurred: one deeply rooted in enterobacteriales and one in a pasteurellales ancestor . In the enterobacteriales branch , besides Escherichia , Shigella , and Salmonella , Lsr organizational homologs were found in Enterobacter , Photorhabdus , and Xenorhadbus species , although most of these instances lacked importer genes ( lsrACDB ) . While it is thought that regulatory proteins conserved across such long phylogenetic distances often regulate different targets [29] , the regulation of community-related functions by different manifestations of the Lsr system ( such as biofilm maturation checkpoints in E . coli [25] and possible biofilm dispersion in B . cereus [12] ) suggests a convergent tendency to leverage a quorum/environment sensing capacity inherent to the Lsr system . Indirect influence over a broader regulon may be abetted by the involvement of AI-2 , the Lsr system substrate , in metabolic pathways [30] .
LMNAST is a program that evaluates similarity or homology on the level of gene organization , conducting a search for patterns and prevalence constrained by a BLAST E-value filter . Program results overlaid onto phylogenetic data allow visual inspection of phylogenetic density and dispersion . 2D homology plots display system variability among LMNAST orthologs , and when overlaid with genera/species clustering , reveal the degree of system conservation within and across genera/species when organizational homology decreases and element homology is constant . Clustering also enables the identification of conserved system homologs . Organizational information is lost when using coincidence heat charts , but suggestions of the underlying structural variability remain nonetheless . This is particularly true for coincidence representations of extended window searches . For such searches , contextual associations with non-canonical genes may also emerge . Trackback plots illustrate both variety and structural information , albeit in a less dense format . These representations are especially useful in combination . It should be noted that the results are almost entirely comprised of excerpts from fully sequenced genomes . Results are also biased by BLAST input , as characters with more element homologs ( e . g . lsrA ) appear more frequently in hits . Generically , LMNAST identified query homologs with a variety of deletions , insertions , misordering , and misdirections . While nearly any source of mutagenesis may result in a frank mutation affecting a system's organizational homology , homologous recombination , insertion sequences , transposable elements , and combinations thereof are likely to be of particular consequence for LMNAST searches . Deletions may be a result of pseudo gene conversion , of chromosomal rearrangements , or part and parcel of an insertion event—if the insertion results in a gap sufficiently large as to disconnect hit elements from one another . In the case of such insertions , sufficiently weak criteria may be of use , with the caveat that decreased stringency increases the number of false positives . From a signaling perspective , depending on the impacted elements and the nature of the inserted sequence , gap presence could result in system discoordination; and the longer the gap the more probable and severe the discoordination , most likely to the detriment of system function . As for the specific test queries examined herein , while the lac operon is well characterized in its canonical form , there nonetheless exists a great deal of frank variation from the textbook case . Of particular interest were homologous instances where structural rearrangement could influence self-regulation of component expression . Also of note were its multiple signaling component deletions . Such abbreviated homologs were frequently repurposed in a related context . Complete lac operons were found among nearly all E . coli strains . Incomplete lac operons were found to be distributed only among closely related Enterobacteriaceae species comprised almost entirely of Escherichia , Citrobacters , Enterobacters , and Serratias as expected based on limited lac operon HGT [18] . This difference between the rates of decay for the two homology signals over phylogenetic space may be suggestive of distinct selection pressures guiding the two systems . Also identified through LMNAST were conserved , E . coli-specific evolved beta-galactosidase systems [17] , demonstrating a capacity to find directly evolved but highly distinct ( meH∼0 . 19 ) homologs . On par , Lsr system hit structural similarity was less well correlated with meH than lac operon results , a phenomenon presumably associated with apparent Lsr system HGT . The Lsr system was phylogenetically dispersed more widely than the lac operon , even while its distribution remained densest among gammaproteobacteria . Much like the lac operon , Lsr system structure was subject to significant variability . lsrK and lsrR characters were common to many hits . lsrF and lsrG were the least common; the inclusion of both elements nearly always coincided with the presence of all other Lsr characters as well . Lsr-contextually associated genes and novel putative Lsr systems were also elucidated . The dispersion of Lsr to bacteria as far afield as the S . smaragdinae , the first Spirochaeta to be fully sequenced [31] , is intriguing . It suggests that while the depth of Lsr dispersion may not be significant , that its exposed breadth will expand incrementally at a rate proportional to microbial genome sequencing . While the direct regulon of such HGT systems is expected to be limited [29] , [32] , the proximity of the substrate to key metabolic pathways may allow the Lsr system to confer contextual phenotypic advantages by impacting downstream pathways with its capacity to recompartmentalize a metabolic intermediate . Moreover , the known regulatory requirements for functional integration of the Lsr system are minimal , consisting entirely of interaction with cAMP-CRP complex , which is deeply rooted in eubacteria . Gene organization differences between dispersed Lsr homologs , may indicate distinct signaling outcomes , in turn suggesting the appropriation of the Lsr system's inherent quorum capacity to drive distinct phenotypes suited to a given bacteria's needs within its particular niche . Unlike the results for the lac operon , Lsr system results returned a large number of other-annotated , low homology systems . This speaks to both the inherent difficulty of extrapolation based on homology and the utility of the additional , complementary homology measure yielded by LMNAST searching . Overall , given the complexity of the results , numerous aspects may be of interest . For example , extant variation of the queried modular systems , as captured by frank changes in gene organization , was revealed . Several topological curiosities were also revealed . For example , the Lsr system's apparent dispersion through both horizontal and vertical inheritance could , in fact , suggest that quorum sensing behavior that is regulated by the Lsr system is conveyed as a root of selective advantage , as opposed to the specific regulon known to uptake small molecules that could otherwise be viewed as carbon source . By considering our results in the context of common graphical tools of a complementary nature ( e . g . 2D similarity plots and coincidence heat maps ) , through LMNAST we offer a new avenue by which to explore this and other provocative questions . | Bacteria communicate with each other through a network of small molecules that are secreted and perceived by nearest neighbors . In a process known as quorum sensing , bacteria communicate their cell density and certain behaviors emerge wherein the population of cells acts as a coordinated community . One small signaling molecule , AI-2 , is synthesized by many bacteria so that in a natural ecosystem comprised of many secreting cells of different species , the molecule may be present in an appreciable concentration . The perception of the signal may be key to unlocking its importance , as some cells may recognize it at lower concentrations than others , etc . We have created a searching algorithm that finds similar gene sets among various bacteria . Here , we looked for signal transduction pathways similar to the one studied in E . coli . We found exact replicas to that of E . coli , but also found pathways with missing genes , added genes of unknown function , as well as different patterns by which the genes may be regulated . We suspect these attributes may play a significant role in determining quorum sensing behaviors . This , in turn , may lead to new discoveries for controlling groups of bacteria and possibly reducing the prevalence of infectious disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"sequence",
"analysis",
"bio-ontologies",
"biology",
"computational",
"biology",
"signaling",
"networks"
] | 2012 | Gene Network Homology in Prokaryotes Using a Similarity Search Approach: Queries of Quorum Sensing Signal Transduction |
Synaptic pathology is an early feature of prion as well as other neurodegenerative diseases . Although the self-templating process by which prions propagate is well established , the mechanisms by which prions cause synaptotoxicity are poorly understood , due largely to the absence of experimentally tractable cell culture models . Here , we report that exposure of cultured hippocampal neurons to PrPSc , the infectious isoform of the prion protein , results in rapid retraction of dendritic spines . This effect is entirely dependent on expression of the cellular prion protein , PrPC , by target neurons , and on the presence of a nine-amino acid , polybasic region at the N-terminus of the PrPC molecule . Both protease-resistant and protease-sensitive forms of PrPSc cause dendritic loss . This system provides new insights into the mechanisms responsible for prion neurotoxicity , and it provides a platform for characterizing different pathogenic forms of PrPSc and testing potential therapeutic agents .
Prion diseases are fatal neurodegenerative conditions of humans and animals that have significantly impacted public health and the safety of the food and blood supplies . These disorders are caused by infectious proteins called prions , which propagate themselves by a self-templating mechanism in which PrPSc , the infectious isoform , seeds conformational conversion of PrPC , a normal neuronal glycoprotein , into additional molecules of PrPSc [1–3] . Although this model for prion infectivity is now widely accepted , the cellular and molecular mechanisms by which prions actually cause neurodegeneration remain a mystery . There is a critical need to address this question in order to develop effective treatments for these currently incurable disorders . The terminal neuropathology of prion diseases encompasses a number of features , including spongiform change , amyloid deposition , astrogliosis and neuronal loss[4] . However , some of the earliest and potentially most critical changes occur at the level of the synapse[5] . Synaptic pathologies , including loss as well as morphological and functional abnormalities of synapses , occur early during the course of many prion diseases , and PrPSc is often found in neuropil deposits that are referred to as “synaptic-like” , since they appear to surround synaptic sites[6–11] . Two-photon imaging studies of living , prion-infected animals demonstrate that swelling of dendritic shafts and retraction of dendritic spines occur early during the disease course , well before symptoms appear[12] . Taken together , these studies pinpoint synapses , in particular dendrites and dendritic spines , as important initial targets of prion neurotoxicity . Dendritic spines are protuberances on dendrites at which synaptic contacts ( usually excitatory ) occur[13] . Changes in their morphology are now believed to underlay synaptic plasticity associated with learning and memory , as well as degenerative events that occur during aging and neurological disease[14 , 15] . A major roadblock in studying prion neurotoxicity has been the lack of an experimentally tractable model system in which degenerative changes can be studied in cell culture . Having a neuronal culture system that is susceptible to the synaptotoxic effects of prions is crucial for delineating the underlying cellular and molecular mechanisms , assaying and characterizing different toxic species of PrPSc , and potentially identifying drugs that block neurodegeneration . There are a limited number of cell lines capable of propagating prions in culture[16 , 17] , and none of these exhibit signs of cytotoxicity as a result of chronic prion infection . Moreover , most of the cells used to propagate prions are transformed cell lines ( e . g . , N2a neuroblastoma cells ) , and there is very little published literature on prion infection of cultured primary neurons [18 , 19] . In this paper , we describe a new system that is capable of reproducing acute prion neurotoxicity , based on PrPSc-induced degeneration of dendritic spines on cultured hippocampal neurons . The effects observed in our system are specific to PrPSc-containing samples , require expression of full-length PrPC by the target neurons , and are apparent within hours , well before chronic infection is established . Using this system , we have made several new observations relevant to prion neurotoxicity .
We monitored the effect of PrPSc-containing brain homogenates on the integrity of dendritic spines displayed by differentiated cultures of hippocampal neurons . Neurons in these cultures , which are maintained in vitro for 3 weeks in the presence of a feeder layer of astrocytes , develop morphologically mature axons and dendrites , and form functional excitatory and inhibitory synapses[20] . The dendrites are studded with a high density of mushroom-shaped and stubby spines ( loci of glutamatergic , excitatory synapses ) , which can be stained with fluorescently labeled phalloidin by virtue of its ability to binds to actin filaments within the spines . After 24 hr of treatment with RML-infected brain homogenate ( IBH ) ( final concentration of 0 . 16% ) the neurons were fixed and stained with Alexa 488-phalloidin , and the number and area of spines were quantitated . The cultures were co-stained for tubulin to reveal the overall morphology of the dendrites . We found that treatment with IBH significantly altered the morphology of dendritic spines ( Fig 1D and 1E ) . There was dramatic retraction of spines , reducing their number per unit length of dendrite , as well as the area of each dendritic spine ( Fig 1K and 1L ) . Spine retraction was accompanied by collapse of the actin cytoskeleton , resulting in residual patches of Alexa 488-phalloidin staining along the dendrite at the former sites of spines ( Fig 1D and 1E ) . Importantly , the microtubule structure of these neurons remained intact in both dendrites and axons , as indicated by tubulin staining ( Fig 1F ) , demonstrating that IBH was causing an early and selective loss of spines prior to major alterations of neuronal morphology or cell death . Normal brain homogenate ( NBH ) from age-matched , uninoculated mice had no detectable effect on dendritic spines ( comparable with untreated cultures: 0 . 78 ± 0 . 03 spines per μm; 81±9 . 4 A . U . average spine area ) , suggesting that the toxicity observed was specific to scrapie-infected homogenate ( Fig 1A , 1B , 1C , 1K and 1L ) . To test whether endogenous PrPC is needed for the toxic effect of IBH on dendritic spines , we treated neurons from PrP knock-out ( Prn-p0/0 ) mice with IBH . In contrast to wild-type neurons , Prn-p0/0 neurons showed no significant change in spine number or area after treatment with IBH ( Fig 1G–1L ) . As an alternative method to visualize dendritic spines , we infected neurons with a lentivirus encoding GFP , which fills the neuronal cytoplasm , including the inside of spines . We found that IBH induced the shrinkage and disappearance of the GFP-labeled dendritic spines , correlating with changes in phalloidin staining ( S1 Fig ) . This procedure made it clear that dendritic spines were actually retracting and disappearing in response to IBH , and that patches of actin were present at the sites of the collapsed spines . We conclude that scrapie-infected brain homogenate causes a rapid , PrPC-dependent retraction of dendritic spines with little effect on overall dendritic morphology . We performed several kinds of experiments to demonstrate that PrPSc is the component of IBH that causes dendritic spine loss , and that other toxic molecules ( e . g . cytokines generated as a result of infection ) are not responsible . First , we took advantage of the fact that the N-terminal domain of PrPC is known to bind specifically to PrPSc , even in a complex mixture of proteins[21–24] . We mixed recombinant PrP 23–110 with IBH prior to treatment of hippocampal neurons , with the expectation that the recombinant protein would bind to PrPSc in the brain homogenate and render it incapable of interacting with PrPC on the neuronal cell surface to produce toxic effects . Consistent with this prediction , we found that addition of PrP 23–110 neutralized the ability of IBH to reduce dendritic spine number and area ( S2 Fig ) . We also tested the effect of two different purified preparations of PrPSc . First , we purified PrPSc from RML-infected mouse brains in the absence of protease treatment . These preparations , which we estimate to be >50% pure ( Fig 2A ) , caused significant loss of dendritic spines and reduction in area of the remaining spines ( Fig 2C , 2F and 2G ) . Using quantitative dot blotting , we estimated that the final concentration of purified PrPSc in the medium used to treat the neurons was 4 . 4 ± 1 . 1 μg/ml , comparable to the final concentration of PrPSc in experiments with crude brain homogenate ( 7 . 5 ± 2 . 4 μg/ml ) . These effects were seen on wild-type neurons , but not on Prn-p0/0 neurons ( Fig 2E–2G ) . Samples prepared from uninoculated brains by the same series of steps had no effect ( Fig 2B , 2D , 2F and 2G ) . We also purified PrPSc using a procedure that involves treatment with pronase E followed by precipitation with sodium phosphotungstic acid ( NaPTA ) in presence of detergent , which efficiently removes PrPC , but leaves PrPSc intact . This procedure results in preparations of higher purity ( >90%; Fig 3A ) because of proteolysis of contaminating proteins . Pronase E digestion preserves protease-sensitive forms of PrPSc , which are typically digested by the proteinase K included in many PrPSc purification methods . We treated neurons with a final concentration of pronase E-purified PrPSc ( ~4 . 4 μg/ml ) that was equivalent to that of the PrPSc purified without protease . Pronase E-purified PrPSc caused significant loss of dendritic spines and reduction in the area of the remaining spines , effects that were seen on wild-type neurons , but not on Prn-p0/0 neurons ( Fig 3C and 3E–3G ) . Samples prepared from uninfected brain by the same series of steps had no effect ( Fig 3B , 3D , 3F and 3G ) . Taken together , these results argue strongly that PrPSc is the component in our brain-derived preparations that is responsible for the toxic effects on hippocampal dendritic spines , and that these effects are dependent on expression of endogenous PrPC by the target neurons . PrPSc purified from brain typically consists of both PK-resistant and PK-sensitive species , which may have different toxic properties . We therefore tested the effect of PK-treated PrPSc on dendritic spine integrity . Based on quantitative Western blotting , PK treatment resulted in digestion of ~90% of the PrPSc , which represents PK-sensitive PrPSc . This proportion of PK-sensitive PrPSc is similar to that reported in other studies[25] . As shown in Fig 4A , treatment with PK resulted in a highly purified preparation of PrPSc ( >95% purity , based on silver staining ) , due to the digestion of non-specific proteins . Treatment with PK also resulted in a downward shift in size due to removal of the N-terminal ~65 amino acids . When we tested the toxicity of PK-digested PrPSc at a concentration of 4 . 4 μg/ml , equivalent to the concentration of non-PK-treated material used in Fig 2 , we observed roughly comparable effects: approximately 85% reduction in spine density and 30% reduction in spine area , with no significant effect on Prn-p0/0 neurons , and no effect of mock-purified material from uninfected brains ( Fig 4B–4G ) . These results indicate that residues 23 through ~90 of PrPSc are not essential for synaptotoxicity . Moreover , since similar toxic effects were seen with equivalent concentrations of undigested and PK-digested PrPSc , our data suggest that both PK-sensitive and PK-resistant forms of PrPSc may contribute to dendritic spine loss . The concentration of purified PrPSc used in the experiments shown in Figs 2–4 was 4 . 4 μg/ml , similar to the estimated concentration of PrPSc in the crude IBH used in Fig 1 . To determine the minimum PrPSc concentration required to observe a toxic effect on dendritic spines , we tested the dose dependence using hippocampal neurons from wild-type mice , as well as Tga20 mice which express ~10X the endogenous level of wild-type PrPC ( S3 Fig ) . At the highest concentration tested ( 4 . 4 μg/ml ) , purified PrPSc reduced spine density by ~85% and spine area by ~30% . Smaller , but statistically significant effects ( 50% and 25% reduction in spine density and area , respectively ) were seen with 2 . 2 μg/ml PrPSc . At 1 . 1 μg/ml PrPSc , there were trends for a reduction in spine density and area , but the differences were not statistically significant . At each concentration of PrPSc , there was no significant difference between WT and Tga20 neurons . As a control , a mock-purified preparation from uninfected brain , equivalent to the highest PrPSc concentration , had no effect on spine density or area . Taken together , these results indicate that synaptotoxic effects of PrPSc can be detected at concentrations as low as 2 . 2 μg/ml , and that boosting the expression of PrPC beyond the endogenous level does not increase the degree of toxicity . Our previous results indicated that PrPC expression by target neurons is essential for dendritic spine loss induced by PrPSc . We went on to analyze which regions of PrPC are critical for this effect . We prepared hippocampal neurons from two lines of transgenic mice that we have previously constructed: Tg ( Δ23–111 ) and Tg ( Δ23–31 ) , which express PrPC harboring deletions of residues 23–111 and 23–31 , respectively . These mice do not express endogenous PrP ( i . e . they have a Prn-p0/0 background ) , and the expression levels of the Δ23–111 and Δ23–31 proteins are , respectively , 7X[26] and 1x[27] the endogenous PrP levels found in WT mice . We found that neurons from both lines of transgenic mice were completely resistant to the toxic effects of purified PrPSc , with dendritic spine density and area similar to those of neurons treated with mock-purified material from uninfected brains ( Fig 5 ) . These results indicate that a small , polybasic region of PrPC ( residues 23–31 ) expressed on target neurons is essential for spine loss induced by exogenously applied PrPSc .
In this study , we have established an experimental system to detect the synaptotoxic effects of PrPSc based on its ability to cause retraction and loss of dendritic spines on cultured hippocampal neurons . We have shown that crude brain homogenates from scrapie-infected mice , as well as three kinds of purified preparations of PrPSc cause dendritic spine retraction , while similar preparations from normal , uninfected brains have no effect . These results argue strongly that PrPSc itself , rather than other toxic molecules present in infected brain are responsible for the dendritic spine loss in our system . We suggest that our system is registering relatively early events in the pathogenic cascade triggered by PrPSc . Retraction of spines is detectable within 24 hours of PrPSc exposure , and occurs prior to other changes in overall dendritic morphology or neuronal cell death . Importantly , the ability of PrPSc to cause dendritic spine loss is entirely dependent on expression of PrPC by target neurons , and on a small , polybasic region at the N-terminus of the PrPC molecule ( residues 23–31; KKRPKPGGW ) . Our system reproduces the earliest changes in dendritic spine morphology that have been observed in the brains of living , scrapie-infected mice by two-photon imaging[12] , as well as in organotypic slice cultures and fixed , brain sections[6–11] . We therefore believe that our experimental approach reveals mechanisms that are directly relevant to pathological processes that occur in vivo during prion diseases . Spine retraction in our system is accompanied by a major collapse of the actin cytoskeleton of the spine , consistent with the dynamic role of actin filaments in maintenance of spine morphology[28] . As shown in this study , our system has already provided several new insights into prion neurotoxicity , and in the future it promises to further illuminate the underlying cellular and molecular mechanisms , thus leading to identification of potential , new therapeutic targets . PrPSc-induced loss of dendritic spines requires the expression of PrPC on the target neurons , most likely because cell-surface PrPC acts as receptor to bind exogenously added PrPSc . This suggestion is consistent with the observation that neuronal expression of membrane-anchored PrPC is necessary for prion-induced neurodegeneration in vivo[29–32] . One possible scenario is that dendritic loss results directly from binding of PrPSc to PrPC , with PrPC itself acting as a toxicity-transducing receptor . An alternative but not mutually exclusive possibility is that cell surface PrPC is first converted to PrPSc ( or some other misfolded form ) which then elicits a toxic signal . The latter possibility would be consistent with a recent report[33] that cell surface PrPC is converted to PrPSc within minutes of contact with exogenously applied PrPSc . Interestingly , we found that Tga20 neurons , which express 10X the endogenous level of PrPC , are no more susceptible than WT neurons to the spine-retracting effects of PrPSc . This suggests that the expression level of PrPC is not the rate-limiting step in this initial pathological process . Our results highlight a critical role for the N-terminal domain of PrPC , particularly polybasic residues 23–31 , in transducing the synaptotoxic effects of PrPSc . Remarkably , neurons expressing exclusively PrPC molecules missing residues 23–31 or 23–111 are completely resistant to dendritic spine loss induced by PrPSc . This result could be attributable to the previously documented role of residues 23–31 in PrPC binding to PrPSc [22–24] . Alternatively , the N-terminal domain may play a direct role in the ability of PrPC or PrPSc to elicit downstream neurotoxic signals . This hypothesis would be consistent with the requirement for the 23–31 region in the neurodegenerative phenotype of transgenic mice expressing certain PrP deletion mutants , as well as for the spontaneous ion channel activity associated with these mutants[34 , 35] . The N-terminal domain has also been shown to be essential for the ability of certain anti-PrP antibodies to elicit neuronal cell death in brain slices [36] . The results presented here shed light on the nature of the toxic species responsible for prion neurotoxicity . PrPSc purified from brain is known to be heterogeneous in terms of aggregation state , protease resistance , and possibly protein conformation [25 , 37–40] . Moreover , there is evidence that infectivity ( the ability to self-propagate ) and neurotoxicity ( the ability to produce neuropathology ) may be distinct properties attributable to different molecular forms of misfolded PrP[41–44] . Although historically PK resistance has been used to define PrPSc in biochemical analyses , it has been estimated that a large fraction of the PrPSc present in the brain after prion infection is actually sensitive to PK digestion [25 , 38–40] . There is debate about the relative infectivity of the PK-resistant and PK-sensitive forms , and it has been suggested that the latter species may represent small aggregates that are particularly neurotoxic without being infectious [41 , 43] . We have demonstrated here that both protease-sensitive and protease-resistant forms of PrPSc have synaptotoxic activity in our assay , and that the N-terminal domain of PrPSc that is removed by PK treatment is not essential for this activity . These results suggest that multiple molecular forms of PrPSc differing in aggregation state and quaternary structure may possess neurotoxic activity . Recently , Aguzzi and colleagues have described an organotypic cerebellar slice system in which scrapie infection results in neuronal death , and they have used this system to study some of the cellular mechanisms underlying prion neurotoxicity[7 , 45 , 46] . This system requires chronic infection of the slices with scrapie , which takes several weeks , and is therefore likely to be registering pathological changes secondary to generalized neuronal loss , in contrast to the earlier , more specific dendritic spine alterations observed in our assay . Our neuronal culture system will make it possible to address several outstanding issues in prion biology . These include identification of the downstream signaling mechanisms that link PrPSc binding by PrPC on the cell surface to neurotoxic sequelae such as dendritic spine retraction , characterization of electrophysiological alterations in synaptic function caused by PrPSc , and analysis of the molecular determinants of neurotoxic vs . infectious forms of PrP . This system may also allow identification of new therapeutic targets , and the testing of compounds that act directly on neurotoxic signaling pathways rather than on the formation of PrPSc . Finally , this system will allow direct comparisons between pathogenic mechanisms involved in prion diseases and other neurodegenerative disorders . Dendritic spine loss is a common theme in many neurodegenerative conditions , including Alzheimer’s , Huntington’s , and Parkinson’s diseases , and has been suggested to contribute to clinical symptoms in patients[47] . Of note , changes in dendritic morphology in cultured hippocampal neurons have been widely used as an experimental readout of the synaptotoxicity of the Alzheimer’s Aβ peptide[48] . Since Aβ oligomers have been shown to bind to PrPC , and to induce dendritic spine retraction that appears analogous to that produced by PrPSc [49–53] , it will be of interest to investigate whether the two neurotoxic aggregates act by similar mechanisms .
Prn-p0/0 mice[54] and Tga20 mice[55] on a C57BL6 background were obtained from the European Mouse Mutant Archive ( EMMA; Rome , Italy ) , and were maintained in a homozygous state by interbreeding . Tg ( Δ23–111 ) mice[26] ( also referred to as Tg ( C1 ) ) and Tg ( Δ23–31 ) mice[27] were constructed as described , and maintained on a Prn-p0/0 background with the transgene array in a hemizygous state . Timed-pregnant C57BL/6 mice ( referred to as wild-type , WT ) were purchased from the Jackson Laboratory ( Bar Harbor , ME ) . Mice were genotyped by PCR analysis of tail DNA prepared using the Puregene DNA Isolation kit ( Gentra Systems ) using primers as described previously[26 , 27] . All procedures involving animals were conducted according to the United States Department of Agriculture Animal Welfare Act and the National Institutes of Health Policy on Humane Care and Use of Laboratory Animals . Hippocampal neurons were cultured from P0 pups as described[20] . Neurons were seeded at 75 cells/mm2 on poly-L-lysine-treated coverslips , and after several hours the coverslips were inverted onto an astrocyte feeder layer and maintained in NB/B27 medium until used . The astrocyte feeder layer was generated using murine neural stem cells , as described[56] . Neurons were kept in culture for 18–21 days prior to PrPSc treatment . All procedures involving animals were conducted according to the United States Department of Agriculture Animal Welfare Act and the National Institutes of Health Policy on Humane Care and Use of Laboratory Animals . Ethical approval ( AN-14997 ) was obtained from Boston University medical center institutional animal care and use committee . Neurons were treated with PrPSc-containing or control preparations for 24 hr , followed by fixation in 4% paraformaldehyde and staining with either Alexa 488-phalloidin or rhodamine-phalloidin ( ThermoFischer Scientific , Waltham , MA ) to visualize dendritic spines , and anti-tubulin antibodies ( Sigma-Aldrich , St . Louis , MO ) to visualize axons and dendrites . Images were acquired using a Zeiss 700 confocal microscope with a 63x objective ( N . A . = 1 . 4 ) . The number and area of dendritic spines were determined using ImageJ software . Briefly , 2–3 dendritic segments with a clear background were chosen from each image , and the images adjusted using a threshold that had been optimized to include the outline of the spines but not non-specific signals[57] . Spine area and number were determined . The number of spines was normalized to the measured length of the dendritic segment to give the number of spines/μm , and the area was normalized to the number of spines to give the average area of each spine in arbitrary units ( A . U . ) . For each experiment , 15–24 neurons from 3–4 individual experiments were imaged and quantitated . A GFP-encoding lentivirus ( created using the vector CSCW-Fluc-IRES-GFP ) was obtained from the Massachusetts General Hospital Vector Core ( https://vectorcore . mgh . harvard . edu ) . Virus ( 105 IU/ml final concentration ) was added to hippocampal neurons after 12 days in culture . Neurons were then cultured for an additional 6 days , to maximize GFP expression , before PrPSc treatment . C57BL6 or FVB mice ( as indicated below ) were inoculated intracerebrally with 30 μl of a 1% brain homogenate from the brain of a corresponding mouse that was terminally ill with RML scrapie . The inoculated mice were monitored until the appearance of clinical signs ( ~170 days post inoculation ) , at which time the animals were euthanized , their brains collected , and stored at -80°C until use . Ten percent ( w/v ) brain homogenates in PBS were prepared from RML infected or non-infected C57BL6 mice using 1 mm glass beads ( D1031-10 , Benchmark Scientific , Edison , NJ ) and a Beadbug microtube homogenizer ( Benchmark Scientific ) . Brain homogenates were passed once through a 0 . 5 cc insulin syringe with a 28 gauge needle ( Becton Dickinson , cat . no . 329461 ) , and aliquoted for storage at -80°C . Two different procedures were used to purify full-length PrPSc from the brains of terminally ill mice infected with the RML strain of scrapie: one that did not use proteases; and another that employed pronase E , which preserves full-length PrPSc . Mock purifications were also carried out from age-match , uninfected brains . The purified samples were evaluated by SDS-PAGE followed by silver staining and Western blotting . Purified PrPSc ( prepared by the protease-free method ) , or the equivalent amount of mock-purified material from uninfected brains , was incubated in a total volume of 250 μl of PBS/2% sarkosyl containing a final concentration of 20 μg/ml of PK ( Roche Diagnostics , cat . no . 03115879001 ) for 1 hr at 37°C . Then 5 μl of 50X Complete Protease Inhibitor Cocktail ( Roche Diagnostics , cat . no . 11836153001 ) was added , followed by 700 μl of PBS . Samples were then centrifuged at 180 , 000g at 4°C for 1 hr . The supernatant was discarded and the pellet resuspended in 21 μl of PBS . 20 μl was used to treat hippocampal neurons , and the remaining 1 μl was used for analysis by Western blotting and silver staining . Proteins in SDS sample buffer were heated at 95°C for 5 min , then resolved by SDS-PAGE in 12% pre-cast gels ( BioRad , cat . no . 567–1044 ) . For Western blotting , proteins were electrophoretically transferred to PVDF membranes ( Millipore , cat . no . IPVH00010 ) . Membranes were blocked for 1 h in 5% ( w/v ) non-fat dry milk in Tris-buffered saline containing 0 . 1% Tween 20 , followed by incubation with anti-PrP antibody D18 ( a human chimeric monoclonal[60] ) , and then with HRP-conjugated anti-human secondary antibody ( Jackson ImmunoResearch , cat . no . 109-035-088 ) . Signals were revealed using HRP substrate ( Millipore , cat . no . WBKLS0500 ) , and were visualized using a BioRad XRS image scanner . Silver staining of gels was carried out using a Silver Stain Kit ( Pierce/ThermoFisher cat . no . 24612 ) following the manufacturer’s instructions . | Prion diseases are fatal neurodegenerative disorders that cause memory loss , impaired coordination , and abnormal movements . The molecular culprit in prion diseases is PrPSc , an infectious isoform of a host-encoded glycoprotein ( PrPC ) that can propagate itself by a self-templating mechanism . Whether PrPSc itself is toxic to neurons , and if so , the cellular mechanisms by which it produces neuronal pathology are largely unknown , in part because of the absence of suitable cell culture models . We describe here a hippocampal neuronal cultural system to detect the toxic effect of PrPSc on dendritic spines , which are postsynaptic elements responsible for excitatory synaptic transmission , and which are implicated in learning , memory , and the earliest stages of neurodegenerative diseases . We found that purified , exogenously applied PrPSc causes acute retraction of dendritic spines , an effect that is entirely dependent on expression of PrPC by target neurons , and on the on the presence of a nine-amino acid , polybasic region at the N-terminus of the PrPC molecule . Both protease-resistant and protease-sensitive forms of PrPSc cause dendritic retraction . This system provides new insights into the mechanisms responsible for prion neurotoxicity , and it provides a platform for characterizing different pathogenic forms of PrPSc and testing potential therapeutic agents . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"animal",
"diseases",
"enzymes",
"enzymology",
"vertebrates",
"neuroscience",
"animals",
"mammals",
"organisms",
"animal",
"prion",
"diseases",
"gel",
"electrophoresis",
"zoology",
"neuronal",
"dendrites",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"infectious",
"diseases",
"electrophoretic",
"techniques",
"staining",
"zoonoses",
"animal",
"cells",
"electrophoretic",
"staining",
"proteins",
"biochemistry",
"cellular",
"neuroscience",
"neuronal",
"morphology",
"cell",
"biology",
"cats",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"proteases",
"amniotes",
"silver",
"staining",
"prion",
"diseases"
] | 2016 | A Neuronal Culture System to Detect Prion Synaptotoxicity |
Dengue-related illness is a leading cause of hospitalization and death in Thailand and other Southeast Asian countries , imposing a major economic burden on households , health systems , and governments . This study aims to assess the economic impact of hospitalized dengue cases on households in Chachoengsao province in eastern Thailand . We conducted a prospective cost-of-illness study of hospitalized pediatric and adult dengue patients at three public hospitals . We examined all hospitalized dengue cases regardless of disease severity . Patients or their legal guardians were interviewed using a standard questionnaire to determine household-level medical and non-medical expenditures and income losses during the illness episode . Between March and September 2015 , we recruited a total of 224 hospitalized patients ( <5 years , 4%; 5–14 years , 20% , 15–24 years , 36% , 25–34 years , 15%; 35–44 years , 10%; 45+ years , 12% ) , who were clinically diagnosed with dengue . The total cost of a hospitalized dengue case was higher for adult patients than pediatric patients , and was US$153 . 6 and US$166 . 3 for pediatric DF and DHF patients , respectively , and US$171 . 2 and US$226 . 1 for adult DF and DHF patients , respectively . The financial burden on households increased with the severity of dengue illness . Although 74% of the households reported that the patient received free medical care , hospitalized dengue illness cost approximately 19–23% of the monthly household income . These results indicated that dengue imposed a substantial financial burden on households in Thailand where a great majority of the population was covered by the Universal Coverage Scheme for health care .
Dengue , an arbovirus infection with an explosive epidemic potential , is a major public health problem in many tropical and subtropical countries today [1] . The incidence of dengue has dramatically increased over the past decades , with the number of symptomatic dengue infections reported to be doubling every 10 years between 1990 and 2013 [2] . Dengue hemorrhagic fever ( DHF ) , a severe and potentially life-threatening form of the disease , has also reported to be increasing steadily during this period , driving the increase in hospitalization rates for dengue , particularly in children [3–7] . Yet , an increasing number of studies have shown that there is substantial under-reporting of dengue cases to national surveillance systems , which prevents an accurate estimation of the disease burden of dengue in endemic countries [8] . Despite its limited effectiveness and high cost , vector control is the mainstay of dengue control and outbreak response in endemic areas [9–12] . The disease is expected to further expand its geographical range due to favorable conditions provided by rapidly growing high density urban areas along with socio-economic changes [13] , increased worldwide travel and trade [14] , and climate change [15] . A growing literature shows that dengue imposes an enormous socioeconomic burden on households , health care systems , and governments in endemic countries [16–18] , particularly during outbreaks [19–22] . Reported as a public health problem since the 1950s , dengue causes frequent outbreaks in Thailand and is hyperendemic with all four distinct serotypes of the dengue virus in circulation for more than five decades [23] . Although dengue has traditionally affected children , there has been a shift in the mean age of dengue cases towards older age groups in Thailand and other dengue hyperendemic countries in Southeast Asia [24–28] . Most dengue cases now occur in individuals aged 5–24 years [29] , which account for one third of the total population in Thailand , and the disease is more common in adolescents and young adults [30] . The incidence of DHF varies widely from year to year , exhibiting as much as a tenfold difference between years [26] . During the period 2000–2011 , the incidence of DHF was higher in children aged 5–14 years than those aged 15 years or older [29] . While the case fatality rate of dengue has been declining steadily over the past decade , the highest rates are seen in children aged 0–4 years [29] . Frequent and severe illness can cause considerable social and economic disruption to households by requiring one or multiple visits to health care providers and hospitalization . Dengue illness often leads to school and work absenteeism , medical and non-medical expenditures , and foregone income . Illness related costs incurred by patients and household members constitute a severe economic burden for households , particularly in developing country settings . To accurately assess the overall economic burden of dengue , cost-of-illness data at the household level are , therefore , essential . Within the context of a European Union funded research project on dengue [31] , we conducted a prospective hospital-based cost-of-illness study to assess the cost and impact of hospitalized dengue cases on households in a highly endemic area in eastern Thailand . Previous cost-of-illness studies in Thailand focused primarily on pediatric dengue patients ( aged under 15 years ) [32–36] . In view of the shift in the age distribution of dengue cases , we expanded the focus to cover adult dengue patients ( aged 15 years and above ) .
We conducted a prospective , hospital-based cost-of-illness study in Chacheongsao province in eastern Thailand . Chachoengsao is a highly endemic area for dengue with a population of 700 , 902 in 2015 [37] and a surface area of 5 , 351 km2 . The province is divided into 11 districts with 93 sub-districts , and has one province-level and nine district-level hospitals in total . Historically an agriculture-based province with rice paddies , fruit plantations , and livestock , it has become industrialized in recent years , transitioning from rural to semi-rural and semi-rural to semi-urban . One provincial-level and two district-level public hospitals participated in the study . The study population included hospitalized pediatric ( aged under 15 years ) and adult ( aged 15 years and above ) patients who were clinically diagnosed with dengue . All patients or their legal guardians were invited to participate in the study and asked to sign an informed consent form . Patients who did not give consent were excluded from the study . The recruitment period was from March to September in 2015 and overlapped with the peak season of dengue illness . We adapted a patient questionnaire , which was successfully used in previous cost-of-illness studies in several dengue endemic countries [17] . It was translated into and back-translated from Thai by two researchers who were fluent in both languages , and the discrepancies were resolved through discussion . The questionnaire was piloted on 10 patients and validated before its administration . It collected information on the demographic and socio-economic characteristics of the patients and other household members , the characteristics of dengue illness episodes , work and school absenteeism , health care service utilization , household health care spending and coping strategies , care provided to the patients by household members , and household income loss due to the dengue illness episode . Patients or their legal guardians were interviewed in-person after recovery from the illness by six experienced public health officers , and each officer received a half-day one-on-one training about the study protocol . The interviews took place at the hospital , the patient’s workplace or home , or any other place convenient for the patient . Each interview lasted about 30–45 minutes . Patients or their legal guardians were compensated for their time with a stipend in the amount of 200 Thai Baht ( THB ) ( US$5 . 8 ) . We followed up with 5–10 patients by phone because there was missing data or inconsistent information in the completed questionnaires . Data were entered into a Microsoft Access Database ( 2015 , Microsoft Corp , Redmond , WA ) and analyzed using SAS 9 . 4 ( SAS Institute Inc . , Cary , NC ) . The unit of analysis was a dengue case , defined as a documented acute febrile illness with a clinical diagnosis of dengue at the time of hospital discharge . This study examined all hospitalized dengue cases regardless of disease severity . Household expenditures on dengue include direct medical and non-medical costs and indirect costs incurred by the household during the dengue illness episode . Direct medical costs comprised all household out-of-pocket payments for medical services received by the patient prior to and during hospitalization . Direct non-medical costs included out-of-pocket payments for transportation , food and lodging for the patient and accompanying household members while seeking and receiving medical care for the illness episode . Indirect costs incurred by the household were assessed as the sum of lost paid work by the patient and other household members aged 15 years and above while caring for the patient during the dengue illness episode . We valued lost paid work as the higher of the reported income loss or the estimated income loss calculated conservatively as the product of the minimum daily wage ( 300 THB [38] ) in Thailand times the number of reported workdays lost by the patient or other household members . The value of time forgone from leisure or other non-market activities was not included in the calculation of indirect costs . If reimbursements were paid to the household by health and/or income protection insurance , the amount reported was subtracted from the sum of the direct and indirect costs for that particular household to arrive at a total cost per case . We also reported on the number of school days lost by the patient and other household members due to dengue illness , as well as the total number of days household members cared for the patient during the illness episode . All costs were presented as mean ( ± standard deviation , SD ) and expressed in 2015 US$ based on 34 . 2 THB to 1US$ currency exchange rate [39] . The protocol for this study was reviewed and approved by the Ethical Review Boards of Mahidol University , Heidelberg University , Chachoengsao Provincial Public Health Office , and Buddhasothorn Hospital . Signed informed consent was obtained from all patients or their legal guardians . Participant information , such as gender , age , clinical diagnostic status and contact information , was obtained from the hospital records . All the data collected through the cost of illness questionnaire and the hospital records were analyzed anonymously .
A total of 570 hospitalized patients , who were clinically diagnosed with dengue , were eligible to participate in the study . Of these hospitalized patients , 224 were recruited into the study . The general characteristics of hospitalized dengue patients and dengue illness episodes are summarized in Tables 1 and 2 . Overall , 48% were female , and 24% were aged under 15 years of age . The age distribution of the study population is presented in Fig 1 . The mean household size was 4 . 4 persons ( SD 1 . 9 ) . Among adult hospitalized patients , 18% had primary school education or less , 36% secondary school education , and 46% vocational/high school/college education or above . Of the 224 hospitalized patients , 168 ( 75% ) and 56 ( 25% ) had a clinical diagnosis of Dengue Fever ( DF ) and Dengue Hemorrhagic Fever ( DHF ) , respectively , at the time of hospital discharge . About 73% of DF patients and 86% of DHF patients were aged 15 years or above . There were no deaths in this cohort of hospitalized dengue patients . Overall , hospitalized dengue patients reported 9 . 5 days ( SD 3 . 4 ) of illness , including 3 . 8 days ( SD 2 . 2 ) during which the patient felt bad or very bad . The mean duration of illness was 8 . 0 days ( SD 2 . 1 ) for pediatric DF patients and 10 . 0 days ( SD 2 . 9 ) for pediatric DHF patients , including 4 . 3 days ( SD 2 . 6 ) and 2 . 5 days ( SD 0 . 8 ) during which the patient felt bad or very bad , respectively . Adult DF and DHF patients , respectively , reported 9 . 5 days ( SD 2 . 5 ) and 10 . 5 days ( SD 4 . 8 ) of illness , including 3 . 5 days ( SD 2 . 1 ) and 4 . 5 days ( SD 2 . 2 ) during which the patient felt bad or very bad . Forty-seven percent of caretakers reported seeking care for their children within 24 hours of onset of illness , 15% reported seeking care one to two days after onset of illness , and 36% waited more than two days . Thirty-six percent adult patients sought care within 24 hours after onset of illness , 27% sought care one to two days after onset of illness , and the remaining 35% waited more than two days . Table 3 presents the type of health facility visited and the type of health provider consulted by hospitalized dengue patients during their illness episode . Fifty-four percent of the patients sought care at a hospital first and got hospitalized during their first visit , followed by 15% visiting a doctor’s office and 14% a pharmacy . Sixty-eight percent of the first visits occurred in a public health facility . About 51% and 17% of the patients reported a second and a third visit , respectively , where 64% and 84% of these visits resulted in hospitalization , and 82% and 97% of them occurred in public health facilities . Two patients had multiple hospitalizations during their illness episode . Dengue patients spent , on average , 3 . 9 ( SD 1 . 9 ) nights in the hospital . While the mean number of hospital nights for pediatric DF and DHF patients was 4 . 0 ( SD 2 . 6 ) and 4 . 6 ( SD 1 . 3 ) , respectively , adult DF and DHF patients spent , on average , 3 . 7 ( SD 1 . 5 ) and 4 . 2 ( SD 2 . 1 ) nights in the hospital , respectively . None of the hospitalized dengue patients reported receiving care in the intensive care unit . Dengue illness affected school attendance and productive activities of the patients and other household members . Table 4 presents the mean number of school days missed and work days lost . Of the 91 hospitalized dengue patients who were studying at the time of illness , 79 reported missing school with an average of 6 . 8 ( SD 4 . 0 ) days . Of the 53 pediatric dengue patients , 48 were in school at the time of illness , and 45 missed an average of 6 . 5 ( SD 3 . 8 ) days of school . The mean number of school days missed was 6 . 6 ( SD 3 . 9 ) and 6 . 1 ( SD 3 . 1 ) days for pediatric DF and DHF patients , respectively . Of the 171 hospitalized adult patients , 43 were in school at the time of illness , and 34 reported missing school with an average of 7 . 2 ( SD 4 . 2 ) days . Of the 99 hospitalized dengue patients who were working for pay at the time of illness , 97 were adult patients , and 94 lost an average of 6 . 9 ( SD 3 . 5 ) days of work due to the illness episode . The mean number of work days lost for adult DF and DHF patients was 6 . 6 ( SD 3 . 6 ) and 7 . 6 ( SD 3 . 1 ) days , respectively . The burden of a hospitalized dengue case on household members was also considerable . Of the 602 household members , 52% and 17% reported to be working and attending school , respectively . On average , household members missed 1 . 2 ( SD 2 . 9 ) days of school and lost 4 . 1 ( SD 3 . 9 ) days of work . The mean total number of days cared for the patient during the illness episode was 7 . 2 ( SD 4 . 9 ) per household . Table 5 presents the direct , indirect and total costs of hospitalized dengue cases to households by patient age category and disease severity . The mean total household cost of a hospitalized pediatric and adult dengue case was US$155 . 4 ( SD 112 . 1 ) and US$186 . 8 ( SD 184 . 7 ) , including a mean reimbursement of US$7 . 7 ( SD 24 . 1 ) and US$20 ( SD 138 . 9 ) , respectively . The direct costs for pediatric and adult patients amounted to US$81 . 9 ( SD 76 . 5 ) and US$109 . 3 ( SD 190 . 4 ) , constituting 52% and 59% of the total household costs while the mean indirect costs were US$81 . 1 ( SD 66 . 3 ) and US$97 . 5 ( SD 110 . 3 ) , respectively . The direct non-medical costs accounted for the majority of the direct costs to households regardless of patient age category , and were US$67 . 2 ( SD 66 . 4 ) and US$78 . 6 ( SD 94 . 7 ) , constituting 82% and 72% of the direct costs , respectively , the rest being the direct medical costs . Overall , the total household cost of a hospitalized dengue case increased with disease severity . The mean total cost of a pediatric DF and DHF case to households was US$153 . 6 ( SD 115 . 3 ) and US$166 . 3 ( SD 96 . 3 ) , respectively . Adult patients reported a mean household cost of US$171 . 2 ( SD 167 . 2 ) and US$226 . 1 ( SD 220 . 1 ) for a DF and DHF case , respectively . The direct non-medical costs similarly accounted for the majority of the direct costs to households regardless of dengue disease severity . Among adult hospitalized patients , the direct medical costs and the indirect costs were notably higher for DHF cases compared to DF cases . Overall , 74% of the households reported that the patient received free medical care during their illness episode , and 62% reported that the patient was covered by the Thai Universal Coverage Scheme for health care . A great majority of the households ( 94% ) reported not borrowing money from outside the household or selling or transfering any household assets to finance the dengue illness episode . About 27% reported using household savings , and 41% reported that other household members helped finance the dengue illness episode .
This study showed that dengue related illness imposes financial hardship on households in Thailand when hospitalization is required . Although direct medical costs were covered for a majority of hospitalized patients by the Thai Universal Coverage Scheme for health care , direct non-medical and indirect costs were of great economic significance to households . These hidden costs of dengue illness are likely to increase given the shift in the mean age and severity of dengue cases in Thailand and other dengue affected countries in the region . This begets the question of whether households can be protected from these hidden costs through innovative policy measures in dengue endemic countries . To fully understand the economic impact of dengue illness on households , it is necessary to collect cost of illness data for both hospitalized and non-hospitalized dengue cases and in both the public and private health sectors . The total cost of a hospitalized dengue case in public facilities accounted for about 19–23% the monthly household income . High household costs of dengue illness strongly justify efforts to improve the coverage of preventive and control measures against dengue . Such cost of illness data are also key to evaluating the cost-effectiveness of these measures , including dengue vaccines . | Dengue , an arbovirus infection with an explosive epidemic potential , is a major public health problem in Thailand and other developing countries in subtropical and tropical regions . Dengue illness often leads to school and work absenteeism , medical and non-medical expenditures , and foregone income . A growing literature shows that these illness related costs pose a severe economic burden on households , health care systems , and governments . We conducted a prospective cost-of-illness study to assess the costs and impact of hospitalized pediatric and adult dengue cases on households in a highly endemic area in eastern Thailand . We found that the total cost of a hospitalized dengue case accounted for about 19–23% the monthly household income . Although direct medical costs were covered for a majority of hospitalized dengue patients by the Thai Universal Coverage Scheme for health care , direct non-medical and indirect costs were of great economic significance to households . These hidden costs of dengue illness are likely to increase given the shift in the mean age of dengue cases in Thailand and other endemic countries in the region . High household costs of dengue illness justify efforts to improve the coverage of preventive and control measures against dengue in endemic areas . | [
"Abstract",
"Introduction",
"Material",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"education",
"sociology",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"indirect",
"costs",
"health",
"care",
"pediatrics",
"neglected",
"tropical",
"diseases",
"patients",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"health",
"economics",
"dengue",
"fever",
"schools",
"economics",
"people",
"and",
"places",
"finance",
"asia",
"viral",
"diseases",
"financial",
"management",
"thailand"
] | 2017 | Household costs of hospitalized dengue illness in semi-rural Thailand |
We provide a novel method , DRISEE ( duplicate read inferred sequencing error estimation ) , to assess sequencing quality ( alternatively referred to as “noise” or “error” ) within and/or between sequencing samples . DRISEE provides positional error estimates that can be used to inform read trimming within a sample . It also provides global ( whole sample ) error estimates that can be used to identify samples with high or varying levels of sequencing error that may confound downstream analyses , particularly in the case of studies that utilize data from multiple sequencing samples . For shotgun metagenomic data , we believe that DRISEE provides estimates of sequencing error that are more accurate and less constrained by technical limitations than existing methods that rely on reference genomes or the use of scores ( e . g . Phred ) . Here , DRISEE is applied to ( non amplicon ) data sets from both the 454 and Illumina platforms . The DRISEE error estimate is obtained by analyzing sets of artifactual duplicate reads ( ADRs ) , a known by-product of both sequencing platforms . We present DRISEE as an open-source , platform-independent method to assess sequencing error in shotgun metagenomic data , and utilize it to discover previously uncharacterized error in de novo sequence data from the 454 and Illumina sequencing platforms .
Accurate quantification of sequencing error is the single most essential consideration of sequence-dependent biological investigations . While true of all investigations that utilize sequencing data , this is particularly true with respect to metagenomics . Metagenomic studies produce biological inferences as the near-exclusive product of computational analyses of high throughput sequence data that attempt to classify the taxonomic ( through 16s ribosomal amplicon sequencing [MG-RAST [1] , QIIME [2]] ) and functional ( through whole genome shotgun sequencing [MG-RAST [1]] ) content of entire microbial communities . The accuracy of these inferences rests largely on the fidelity of sequence data , and consequently , on the ability of existing methods to quantify and account for sequencing error . Surprisingly , the most widespread methods to determine sequencing-error in metagenomic data lack essential features and/or produce underestimates of the overall error that disregard a substantial portion of sequencing-related experimental procedures . Sequence-based experimental inferences , particularly those related to the identification and characterization of features ( protein or 16s rRNA coding regions , regulatory elements , etc . ) are greatly affected by the presence of sequencing errors [3] . Errors in metagenomic amplicon-based sequencing have led to grossly inflated estimates of taxonomic diversity [4] , [5] , [6] . While methods such as denoising [2] , [7] , [8] have been developed to address these issues in amplicon-based metagenomic sequencing [2] , [9] , no analogous techniques have been reported to account for noise/error in the context of shotgun-based metagenomic sequencing . Limitations inherent to methods used to assess de novo sequencing error are largely to blame . At present , two methods are commonly used: reference-genome and score -based . Reference-genome-based methods compare de novo sequenced reads to preexisting standards ( published genomes ) . Samples are typically cultured from a clonal isolate for which a reliable reference genome is readily available . Sequenced reads undergo an initial alignment to the selected reference genome to match de novo sequences with the regions in the reference genome to which they correspond . Reads that do not exhibit a high enough level of identity with the reference genome are excluded from further analysis . Reads that exhibit a high fidelity match to a region in the reference genome are compared to that region in great detail . Deviations between sequenced reads and their corresponding loci in the reference genome are scored as errors; these are typically reported with respect to frequency and type ( i . e . insertion , deletion , substitution ) . Selection of the most appropriate reference genome is essential . This is problematic when the best available reference is a related strain or species . In these cases , real biological variation can be mistaken for sequencing error [10] , [11] , [12] , [13] . Reference-genome-based methods provide a particularly effective means to examine sequencing error in the context of genomic ( i . e . single genome sequencing or re-sequencing ) data , but are not applicable to metagenomic samples as these typically contain enormous taxonomic diversity ( samples contain a broad spectrum of species ) for which little adequate reference data is available . Many species have no appropriate reference genome ( s ) , and reference metagenomes do not currently exist . Score-based methods use an alternative approach . Sequencer signals are compared with sophisticated , frequently proprietary , probabilistic models that attempt to account for platform-dependent artifacts , generating base calls , each with an affiliated quality ( Phred or Q score ) that provides an estimate of error frequency , but no information regarding error type . Although score-based methods are applicable to metagenomic data , their inability to consider error type can prove to be a substantial limitation . For example , similarity-based gene annotation is extremely sensitive to frame-shifting insertion/deletion errors but only moderately affected by substitutions [3] . In this context , knowledge of error type , specifically the ratio of insertion and/or deletions to substitutions provides crucial information , knowledge unattainable with conventional Phred or Q scores . The absence of information regarding error type is an even greater concern in light of documented platform-dependent biases in sequencing error type: Illumina-based sequencing exhibits high substitution rates [14] , whereas 454 technologies exhibit a preponderance of insertion/deletion errors [13]; identical Q scores from these two technologies are likely to represent different types of error , rendering ostensibly similar metrics incomparable [13] , [15] , [16] , [17] , [18] , [19] , [20] . The most concerning , but paradoxically least discussed and perhaps least understood , deficit of score-based methods is their implicit disregard of experimental procedure . Typical sequencing efforts employ a host of procedures to extract , amplify , and purify genetic material , experimental processes that necessarily contribute errors ( i . e . introduction of non-biological bias in sequence content and/or abundance relative to original biological template sequences ) ; however , as these errors are introduced before the actual act of sequencing , they can not be accounted for with score-based methods . Thus , a large portion of experimental error in sequencing is frequently overlooked ( Figure 1a ) ( an in depth literature search revealed no works that directly address this issue ) . Reference-genome and score-based sequencing error determination methods require extensive prior knowledge in the form of reference genomes and/or elaborate platform dependent error models . At present it is not possible to apply reference-genome-based methods to metagenomic data . Score-based methods provide , at best , an incomplete assessment of error that is incomparable between technologies and provides no information with respect to error type . Neither of these approaches is well suited to platform-independent analysis of error in shotgun-based metagenomic data . The absence of an appropriate means to assess sequencing error , in a platform independent manner , in the context of metagenomic data , grows more acute with the increasing democratization of high-throughput sequencing technologies ( www . technologyreview . com/biomedicine/26850/ ) and the rapid proliferation of projects that utilize them [21] , [22] , [23] , [24] ( in addition , www . 1000genomes . org , www . commonfund . nih . gov/hmp , www . earthmicrobiome . org ) . This includes an increasing trend toward meta-analyses ( studies that consider data from multiple sources ) to examine collections of samples that can exhibit a diverse technical provenance [25] , . Meaningful comparisons of technically diverse sequence data require accurate and platform-independent measures of sequencing error , such that bona fide observations can be differentiated from background noise . Current methods , score-based methods in particular , are not well equipped to provide these comparisons .
The limitations of reference-genome and score -based methods inspired the creation of Duplicate Read Inferred Sequencing Error Estimation ( DRISEE ) . DRISEE exploits artifactually duplicated reads ( ADRs ) , nearly identical reads that share a completely identical prefix , present with abundances that greatly exceed chance expectations , even when a modest level of biological duplication is taken into account [12] , [26] . We exploit the highly improbable abundances of ADRs to distinguish them from other reads ( see Methods for details ) . While 100% identity in the prefix region is used to cluster reads , only the non-prefix bases ( those not required to exhibit identity with other reads ) are used in the error calculations . No additional requirement for sequence identity/similarity is required of the non-prefix bases . Given their technical origins , it is reasonable to assume that sequence variation within groups of ADRs are more likely to be the product of technical artifact ( s ) ( i . e . sample processing and/or sequencer errors ) than a reflection of genuine diversity in the originally sampled population or culture . Based on this premise , DRISEE utilizes multiple alignment ( by default , multiple alignments are processed with QIIME [2] integrated Uclust [29] – users will soon be able to choose from a variety of other multiple alignment algorithms ) of groups of prefix-identical clusters of ADRs to create internal standards ( consensus sequences ) to which each individual duplicate read is compared . Sequencing error is determined as a function of the variation that exists within clusters of ADRs . This strategy is platform-independent and can be used to quantify error in metagenomic or genomic samples with respect to error frequency and type . DRISEE identifies duplicate reads using stringent requirements for prefix length and abundance that are extremely unlikely to occur unless the sequences have been artifactually duplicated . In the work presented here , a prefix length of 50 bases and a minimum abundance of 20 reads was used; chance occurrence ≈ 4E-32 ( see Methods ) . It is important to note that this probability is so small as to be deemed effectively impossible in biological sequence data ( by way of comparison , the number of atoms in the human body has been estimated at ∼E28 [30] ) ; however , ADRs routinely exhibit abundances that greatly exceed these expectations , making it relatively easy to identify these sequences and simultaneously differentiate them from much lower abundance biological duplication ( there are obvious exceptions to this notion , conserved regions in 16s ribosomal genes , repetitions in eukaryotic DNA etc . ) . Figure 1b provides a visual overview of DRISEE; text S1 ( Supplemental Methods ) outlines a typical DRISEE workflow in much greater detail . The initial output of a DRISEE analysis is a table , excerpted examples of which are presented as Tables 1 and 2 . It indicates the number ( Table 1 ) , or percent ( Table 2 ) , of sequences ( indexed by consensus sequence position ) in all considered clusters of ADRs that match or do not match the consensus derived from the ADR cluster to which they belong . DRISEE tables can indicate the match/mismatch counts for a single cluster of prefix-identical reads from a single sequencing sample , for multiple clusters from a single sample ( Tables 1 and 2 present one such example ) , or for multiple clusters collected from a large number of samples that may represent some common trait of interest ( e . g . samples produced with the same sequencing technology , that used the same RNA/DNA extraction procedures , that were collected as part of the same sequencing project etc . ) . This adaptable tabular format represents the simplest incarnation of a DRISEE error profile; it can be analyzed and visualized in a number of ways ( numerous examples are presented below – see Figures 2–5 ) to garner detailed platform-independent information regarding sequencing error in genomic and metagenomic shotgun sequencing data . A more detailed description of the tabular format is included in the legend for Tables 1 and 2 . Initial validations of DRISEE with simulated data showed nearly perfect correlations between known and DRISEE-based error estimates ( Figure 2a , R2 = 0 . 99 ) . Additional validations with real genomic sequencing data exhibit good correlation with error estimates produced by conventional reference-genome-based analyses [12] of the same samples ( Figure 2b , R2 = 0 . 89 , excluding outliers ) . In further trials , DRISEE was applied to genomic and metagenomic shotgun data produced by two widely utilized sequencing technologies , 454 and Illumina ( n = 242 genomic 454 , n = 65 metagenomic 454 , n = 10 genomic Illumina , and n = 159 metagenomic Illumina samples ) , 476 samples in all . Less than half of the individual samples ( n = 169 ) exhibit DRISEE-based errors consistent with the reported range of second-generation sequencing errors ( 0 . 25–4% ) [4] , [11] , [12] , [13] , [19] , [31] . The majority of samples ( n = 307 ) exhibit DRISEE-based errors that fall outside the range of reported sequencing errors ( error<0 . 25% , n = 73; error>4% , n = 234; avg ± stdev = 12 . 63±15 . 12 ) ( Figure 3 ) . The Supplemental Methods ( Text S1 ) include a description as to how data sets were selected . To compare DRISEE derived errors with those determined with a more conventional score-based approach , we obtained FASTQ data ( i . e . Phred scores ) via SRA ( http://www . ncbi . nlm . nih . gov/sra ) for subsets of DRISEE-analyzed samples: 20 of the 65 metagenomic 454 samples and 12 of the 159 Illumina metagenomic samples . Per base DRISEE and Phred [32]-based errors for these samples were calculated and compared ( see Methods ) . In 454 and Illumina-based metagenomic sequencing data , DRISEE profiles reveal error levels much higher than those reported by archived Phred values ( Figure 4a & b ) . It is also intriguing to note that , whereas Phred values exhibit nearly indistinguishable trends between the 454 and Illumina data , DRISEE error profiles differ markedly for each technology ( Figure 4a & b ) . After observing differences in error profiles between 454 and Illumina technologies , we explored the possibility that DRISEE could be used to observe differences in sequencing error produced by a single sequencing platform ( Illumina ) . Sequencing samples from five projects ( i . e . groups of samples that were produced together in a single experimental framework ) were explored by comparing the total DRISEE error profile for each ( Figure 4c ) . While two projects exhibited similar error profiles ( Sample Sets 2 and 5 ) , most were unique . The ability of DRISEE to resolve unique error profiles was tested further by exploring two individual samples taken from the same project/experiment ( Sample Set 3 ) , those that exhibited the highest and lowest average DRISEE errors . Although the two samples were produced on the same sequencing platform as part of the same experimental project , the individual error profiles are drastically different ( Figure 4d ) . The two samples underwent annotation via MG-RAST , a summary of the annotation results for each sample appears , as a pie-chart , imbedded in the plot of the DRISEE profiles . We also used DRISEE to provide data regarding error type . Figure 5 presents all error types together ( total error ) as well as a breakdown of each error type ( A , T , C , and G substitutions and insertion/deletion errors ) observed across metagenomic 454 ( 65 samples ) and Illumina ( 159 samples ) data . The results are consistent with previous observations in genomic shotgun sequencing: Illumina data are dominated by substitution-based errors [14] , whereas 454 data exhibit a majority of insertion/deletion errors [13] ( Figures 5a and 5b ) . No other method provides estimates with respect to error type in metagenomic shotgun data .
DRISEE provides a more complete estimate of sequencing error than is possible with score-based methods , one that accounts for error introduced at any/all procedural steps in a sequencing protocol – all steps that have the potential to introduce errors ( i . e . deviation from the original biological template sequences ) – from collection of a biological sample to extraction of DNA/RNA , intermediary processing of the extracted material and , finally , sequencing itself ( see Figure 1a ) . Error introduced by processes outside of the actual act of sequencing are ignored by score-based methods , thus it is not surprising that DRISEE derived errors are generally larger than Q/Phred scores , as they account for errors introduced over a much broader scope of experimental procedures , from sample collection , to a wide variety and number of possible intermediary processes , to sequencing itself . An example may help to illustrate the critical importance of this consideration: Amplification is commonly utilized to generate sufficient quantities of material for sequencing from an initial RNA/DNA sample . Here we refer specifically to amplification performed outside of the sequencer/sequencing protocol . Various methods exist – classically variants of the polymerase chain reaction were used , more recent incarnations have adopted isothermic techniques – all depend on high fidelity enzymes ( e . g . Taq or Φ29 DNA polymerase ) , and are experimental processes , prone to experimental error . Even with high fidelity enzymes , amplification products will contain errors ( i . e . deviations from the original biological template ) . Successive amplification ( s ) propagate previous errors and introduce new ones , leading to populations of reads that increasingly diverge from their original biological templates . Amplification products are frequently used as the starting material for a sequencing run , thus the starting material may contain large numbers of unique reads that do not exist in the original biological sample . Score-based methods have no means to distinguish these unique and non-biological reads from the original templates . Scores do provide useful information , the fidelity with which sequencer base calls are made , but these estimates possess no information with respect to the origin of the sequenced read: is the sequence genuine/biological or an error containing artifact of imperfect amplification ? Through the careful examination of prefix-identical reads , DRISEE is able address this question; in the context of shotgun metagenomic data , no other method can . We assert that reference-genome-based error determination methods provide the most complete and accurate measure of sequencing error . This is due to the fact that ( 1 ) such methods consider the entire scope of procedures that accompany a typical sequencing experiment and ( 2 ) they compare raw sequence data to an absolute standard , a reference genome . Score-based metrics ( e . g . Q or Phred ) only consider error introduced by the actual act of sequencing ( ignoring error introduced by any processes that precede actual sequencing – e . g . DNA/RNA extraction , sample amplification and purification , etc . ) and are the product of proprietary black-box software products that can vary considerably among different sequencing technologies . Unfortunately , reference-genome-based methods cannot be applied to metagenomic data ( reference metagenomes do not exist , and are unlikely to anytime in the near future ) . DRISEE can be thought of as a reference-genome-like method , the key difference is that the reference sequences are derived internally from the pool of artifactual duplicate reads , and not from an external reference genome . The similarity between reference-genome and DRISEE derived errors for the same genomic sequencing samples ( Figure 2b ) is not surprising; both methods rely on comparisons to reference standards . Unfortunately , reference-genome-based methods cannot be applied to metagenomic data ( the appropriate reference standards do not exist ) . Reference-genome-based methods possess another potential fault , the utilization of preliminary identity/similarity filters that may lead to artifactual deflation of error estimates . In particular , conventional reference-genome-based methods employ a preliminary similarity search to align sequenced reads to the most similar portion of the selected reference genome . Reads that fail to align to the reference genome with the selected initial level of stringency ( criteria are generally lenient , e . g . 90% identity for the full length of the read [12] ) are discarded from subsequent analysis . In this way , the most error prone reads , those that do not align well to the reference genome , even with lenient criteria , and would contribute significantly to calculated error , are not considered . DRISEE takes a very different approach . Reads are binned based on 100% identity in their prefix region , but no identity/similarity requirement is made of the non-prefix bases . Criteria for prefix length and abundance provide conditions so improbable as to preclude any possibility other than technical duplication . Technical duplicates should be identical to each other , not just in their prefix region , but through the length of the entire read , except for differences introduced by error . While 100% identity in the prefix region is used to cluster reads , only the non-prefix bases ( those not required to exhibit identity/similarity with other reads ) are used in the error calculations . As no additional requirement for sequence identity/similarity is required of the non-prefix bases , DRISEE can provide estimates of error that are less constrained by filters placed in conventional reference-genome based methods . As an example , consider a 100 bp read . Under the reference-genome-based method utilized by Niu et al . ( see Figure 2b ) , 90 bp would be required to perfectly align with a reference genome before error analyses are conducted; thus , the maximum detectable deviation from the reference standard is 10% ( i . e . a maximum of 10% error can be detected ) . Alternatively , DRISEE would cluster the read into a bin of reads with the same 50 bp prefix and would subsequently ignore this prefix to produce an estimate of error solely on the non-prefix bases ( those not required to exhibit identity/similarity with other reads in their respective bin ) . This allows DRISEE to consider errors that span a much broader range ( errors in excess of 50% have been observed – see Figure 3 ) . Given that DRISEE considers the complete scope of procedures implemented in a given sequencing experiment , and score-based methods only provide information with respect to the actual act of sequencing , it is not surprising that DRISEE produces error estimates that are generally higher ( Figures 3 , 4a , & 4b ) . The uniqueness of DRISEE error profiles was unexpected . Distinct error profiles are observed for each of two sequencing technologies , 454 ( Figure 4a ) and Illumina ( Figure 4b ) ; each exhibits a clearly unique error profile , whereas Q-value derived error profiles for the very same samples are indistinguishable from each other . Furthermore , unique profiles were observed when samples processed with the same sequencing technology ( Illumina ) were grouped by experiment , suggesting the presence of platform-independent technological or lab-dependent errors ( Figure 4c ) . Even finer distinctions are observable among the error profiles for single samples taken from the same experiment ( Figure 4d ) . DRISEE provides a means to assess the relative quality of sequencing between technologies ( Figure 4a and b ) , experiments performed on the same platform ( Figure 4c ) , and even between individual samples taken from the same experiment ( Figure 4d ) . The ability of DRISEE to provide a preliminary estimate of sample quality , and indications as to the suitability of a sample for subsequent analyses , is clearly demonstrated in Figure 4d . Two samples from the same experiment exhibit vastly different DRISEE error levels ( 1 vs . 45% average error ) . These values are reflected in the MG-RAST-based annotations of the samples . Nearly 90% of the reads from the high error sample fail MG-RAST quality control procedures; just 4% of the reads are successfully annotated as known proteins . The higher quality data set loses a much smaller portion of its reads to quality control ( 23% ) and has eight times as many reads annotated as known proteins ( 33% ) . In the current age of compute-constrained bioinformatics , the identification and correction/removal of low quality sequence data , from relatively mild procedures like read trimming – DRISEE informed read trimming is currently under development – to more drastic action , including the elimination of entire sequencing samples , is an acute and steadily growing necessity . DRISEE can provide researchers with the ability to identify low quality sequence data before time-consuming and potentially costly analyses are performed . DRISEE also provides researchers with a platform-independent means to assess error among samples , after they have undergone analyses , allowing a quantitative assessment as to the fidelity of analysis-derived inferences . As an example , annotations related to high error samples like that presented in Figure 4d ( purple DRISEE profile ) should be treated with a great deal more skepticism than those derived from a higher quality data set ( e . g . 4d , green DRISEE profile ) . This is especially true when considering samples with subtle differences that may easily be obscured by high levels of sequencing error . Arguably , DRISEE has some limitations . At present , it is not applicable to eukaryotic data , sequences with low complexity , and/or known sequences that may exhibit an unusually high level of biological repetition , particularly amplicon ribosomal RNA data . These types of data are likely to meet DRISEE requirements for prefix length and abundance , but represent real biological variation that could be misinterpreted by DRISEE as sequencing error . Moreover , DRISEE operates on artifactually duplicated reads—an approach that works well with current platforms such as 454 and Illumina but may require procedural modifications ( such as the intentional inclusion of highly abundant sequence standards ) if future developments eliminate ADRs . In summary , DRISEE provides accurate assessments of sequencing error of metagenomic ( Figures 3–5 ) and genomic ( Figure 2 ) data , accounting for error type as well as frequency ( Figure 5 ) . DRISEE error profiles can be used to explore correlations between sequencing error and metadata ( e . g . Figure 4a & b suggests the presence of platform dependent trends in DRISEE calculated errors; Figure 4d demonstrates a correlation between DRISEE calculated error and the percent of reads that MG-RAST is able to successfully characterize ) , allowing investigators to differentiate experimentally meaningful trends from artifacts introduced by previously uncharacterized sequencing error . Traditional score- and reference-genome- based methods do not allow for such observations with respect to shotgun metagenomic data . DRISEE also offers the advantage that it requires no data other than an input FASTA or FASTQ file . Moreover , DRISEE considers error independent of sequencing platform , without prior knowledge . These characteristics make DRISEE a promising method—particularly with respect to the enormous quantities of shotgun-based metagenomic data that are anticipated in the near future . DRISEE will soon be available to analyze sequencing samples in MG-RAST . We also provide MG-RAST independent code to allow users to perform DRISEE analyses without MG-RAST: https://github . com/MG-RAST/DRISEE .
Duplicate Read Inferred Sequencing Error Estimation ( DRISEE ) can be applied to sequence data produced from any sequencing technology . It provides an error profile ( Tables 1 and 2 provide an excerpted example ) that can be used to explore the sequencing error , as well as biases in error , that are present in a single sequencing run or any group of sequencing runs . The latter capability enables the user to produce error profiles specific to a particular sequencing technology , sample preparation procedure , or sequencing facility—in short , to any quantified variable ( i . e . , metadata ) related to one or more sequencing samples . DRISEE exhibits several desirable characteristics that are not found in the most widely utilized methods to quantify sequencing error: reference-genome-based methods that rely on comparison to standard sequences ( generally a published sequenced genome ) : and quality score-based methods that rely on sophisticated , platform-dependent models of error to derive base calls with affiliated confidence estimates ( Q or Phred scores ) for each sequenced base . DRISEE can be applied to metagenomic or genomic data produced with any sequencing technology and requires no prior knowledge ( i . e . , reference genomes or platform–dependent error models ) . DRISEE relies on the occurrence of artifactually duplicated reads—nearly identical sequences that exhibit abundances that greatly exceed expectations of chance , even when a modest amount of possible biological duplication is taken into account . Illumina and 454 platforms exhibit a well documented [12] , [26] , but poorly understood , propensity to produce large numbers of ADRs . DRISEE utilizes this artifact as a means to create internal sequence standards that can be used to assess error within a single sample , or across multiple samples . We identify ADRs as those reads that exhibit an identical prefix ( prefix = the first l bases of a read ) at some threshold abundance ( n ) that exceeds chance expectations , even those that take biological duplication into account . The precise values of l ( prefix length ) and n ( prefix abundance ) can be varied to accommodate the scale of any sequencing technology . In the work presented here , bins ( groups ) of duplicate reads were used to calculate error values if they exhibited an identical prefix length ( l ) of 50 bases with an abundance ( n ) of 20 or more reads . These requirements are arbitrary , but were selected on sound statistical and biological assumptions . Chief among these is the extreme improbability that such an occurrence ( 20 reads , each with identical 50 bp prefixes ) could occur by chance , ( i . e . without technical duplication via WGA or PCR etc . These criteria are stringent enough to justify assumptions of biological and statistical uniqueness; indeed , such an occurrence is extremely unlikely by chance:where p is the probability that a prefix of length l ( 50 bp ) will be observed n ( 20 ) times; 4 represents the number of possible bases ( A , T , C , and G ) . Even in data that are Illumina scale ( on the order of 1 million reads per run ) , a chance observation of 20 reads that exhibit the same 50 bp prefix is highly improbable ( chance≈1E06×4E-32 = 4E-26 ) ; however , ADRs frequently exceed these limits , making them easy to detect , and providing an ideal population to probe for sequencing error – a population of reads that should be completely identical ( i . e . identical beyond their 50 bp prefix ) except for errors introduced by sequencing procedures . The default values for nucleotide length and number of reads required for a bin of ADRs to undergo DRISEE analysis are arbitrary; however , they possesses a key feature , improbability far beyond that expected by chance , even if biological repetition was present , and even when data are Illumina scale ( 1E06 reads ) . Less stringent criteria ( prefix length 20 bp , prefix abundance 20; p = 5E-14 ) were applied to data generated by 454 technologies , yielding extremely similar estimations of error ( data not shown ) . Much more stringent criteria were selected for this study such that the method could be applied to 454 and Illumina data without concern for the difference in scale in the outputs of the two technologies ( 454≈1E05 , Illumina≈1E06 reads per run ) . DRISEE exhibits a universality that other methods lack , but only if the data under consideration meet the following criteria: ( 1 ) Data must be in FASTA or FASTQ format . ( 2 ) There must be enough ADRs to safely infer that they are the product of artifact and not of real biological variation . ( 3 ) Input sequence data should not be culled , trimmed , or modified in any way by sequencer processing software: note that while DRISEE utilizes ADRs in its calculations , it does not cull these sequences from processed datasets ( 4 ) Data under consideration should be the product of random ( i . e . shotgun ) sequencing . ( 5 ) At this time , amplicon data—specifically , directed sequencing of amplicon ribosomal RNA data , are not suitable for DRISEE analysis; ribosomal amplicon reads start with highly conserved regions ( primer target sites ) followed by regions that exhibit a large degree of real biological variation ( the hypervariable regions ) that DRISEE could misinterpret as error . Unless otherwise indicated , data sets examined in this study were obtained via SRA or MG-RAST . Table S1 ( Supplemental Table 1 ) contains a complete list of sequence data used in the accompanying manuscript . Datasets are referenced by their SRA ( http://www . ncbi . nlm . nih . gov/sra ) , MG-RAST ( http://metagenomics . anl . gov/ ) , or both identifiers/accession numbers . An MG-RAST independent version of DRISEE code , with detailed documentation , including installation and running instructions as well as runtime related statistics , can be downloaded from https://github . com/MG-RAST/DRISEE . See Text S1 ( Supplemental Methods ) and Figure 1b for a detailed workflow-based description of DRISEE . DRISEE analysis tables take the same form if they exhibit the counts derived from a single bin of artificially duplicated reads , multiple bins from the same sample , or much larger collections of bins spanning multiple samples . The excerpted tables displayed here represent the raw and percent scaled DRISEE error profile for all considered prefix-identical bins in a single metagenomic sequence sample ( MG-RAST ID 4462612 . 3 ) . The DRISEE table is presented as raw counts per base pair position ( Table 1 ) or percent error per position ( Table 2 ) . Tables 1 and 2 contain three sections ( ID , Summary , and bp counts ) , described in the legends below . | Sequence quality ( referred to alternatively as the level of sequencing error or noise ) is a primary concern to all sequence-dependent investigations . This is particularly true in the field of metagenomics where automated tools ( e . g . annotation pipelines like MG-RAST ) rely on high fidelity sequence data to derive meaningful biological inferences , and is exacerbated by the capacity of next generation sequencing platforms that continue to expand at a rate greater than Moore's law . We demonstrate that the most commonly utilized means to assess sequencing error exhibit severe limitations with respect to analysis of metagenomic data . Furthermore , we introduce a method ( DRISEE ) that accounts for these limitations through the application of a novel approach to assess sequencing error . DRISEE-based analyses reveal previously unobserved levels of sequencing error . DRISEE provides a platform independent measure of sequencing error that objectively assesses the quality of entire sequence samples . This assessment can be used to exclude low quality samples from computationally expensive analyses ( e . g . annotation ) . It can also be used to evaluate the relative fidelity of analyses after they have been performed ( e . g . annotation of error prone samples is less reliable than that of samples with low levels of sequencing error ) . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genome",
"sequencing",
"genomics",
"metagenomics",
"mathematics",
"statistics",
"biology",
"computational",
"biology",
"biostatistics"
] | 2012 | A Platform-Independent Method for Detecting Errors in Metagenomic Sequencing Data: DRISEE |
Nontuberculous mycobacterial pulmonary disease ( NTM-PD ) has become an emerging infectious disease and is responsible for more deaths than tuberculosis in industrialized countries . NTM-PD mortality remains high in some series reportedly ranging from 25% to 40% at five years and often due to unfavorable evolution of NTM-PD despite established treatment . The purpose of our study was to search for early factors that could predict the favorable or unfavorable evolution of NTM-PD at the first year of treatment . In this retrospective and multicenter study , we selected 119 patients based on clinical , radiological and microbiological data from 2002 to 2012 from three French university hospitals ( Guadeloupe , Martinique , Montpellier ) with definite ( meeting the criteria of the American Thoracic Society and the Infectious Disease Society of America in 2007; ATS/IDSA ) or probable ( one positive sputum culture ) NTM-PD . We compared two patient groups: those who improved at one year ( clinical symptoms , radiological lesions and microbiology data ) and those who did not improve at one year . The data were analyzed for all patients as well as for subgroups by gender , HIV-positive patients , and Mycobacterium avium complex ( MAC ) infection . The average patient age was 50 years ± 19 . 4; 58% had respiratory comorbidities , 24% were HIV positive and 19% had cystic fibrosis . Coughing concerned 66% of patients and bronchiectasis concerned 45% . The most frequently isolated NTM were MAC ( 46% ) . 57% ( n = 68 ) of patients met the ATS criteria and improved status concerned 38 . 6% ( n = 46 ) . The improvement factors at one year of NTM-PD were associated with the duration of ethambutol treatment: ( Odds ratio adjusted [ORa]: 2 . 24 , 95% Confidence interval [CI]; 2 . 11–3 . 41 ) , HIV-positive status: ( ORa: 3 . 23 , 95% CI; 1 . 27–8 . 45 ) , and male gender: ( ORa: 2 . 34 , 95% CI; 1 . 26–8 . 16 ) . For the group with NTM-PD due to MAC , improvement was associated with the duration of macrolide treatment ( ORa: 3 . 27 , 95% CI; 1 . 88–7 . 30 ) and an age <50 years ( ORa: 1 . 88 , 95% CI; 1 . 55–8 . 50 ) . In this retrospective multicenter study , improvement at one year in patients with definite or probable NTM-PD was associated with the duration of ethambutol treatment , HIV-positive status and male gender . For the group of patients infected with MAC , improvement was associated with the duration of macrolide treatment and an age <50 years . Identifying predictors of improvement at one year of NTM-PD is expected to optimize the management of the disease in its early stages .
Infection with nontuberculous mycobacteria ( NTM ) preferentially affects the lungs and occurs by inhalation of aerosols containing mycobacteria [1 , 2] . NTM are ubiquitous environmental bacteria found in soil , but also in other sources such as contaminated water taps . The frequency of NTM species can vary from region to region in the world [1 , 3] . NTM pulmonary disease ( NTM-PD ) has today become an emerging infectious disease in industrialized countries . Its increasing prevalence is estimated at more than 50 cases per 100 , 000 persons in some demographic groups in the US [4]; while its incidence in Europe ranges from 0 . 2 to 2 . 9 / 100 , 000 inhabitants [1] . Remarkably , all NTM species are not likely to cause NTM-PD; only a few species such as Mycobacterium avium complex ( MAC ) , M . abscessus , M . xenopi and M . kansasii are frequently involved [5] . Indeed , the clinical relevance of NTM differs by species since they are not endowed with the same virulence [6] . The diagnostic criteria of the American Thoracic Society and the Infectious Disease Society of America in 2007 ( ATS/IDSA ) [7] have established the diagnosis of NTM-PD based on clinical symptoms , radiological lesions and microbiology data . During this decade , real progress has been made in the understanding of this disease [4] . We know for example that besides immunosuppression by HIV or cystic fibrosis , NTM-PD occurs in lungs whose architecture is already weakened by chronic respiratory diseases such as primarily chronic obstructive pulmonary disease ( COPD ) and bronchiectasis [1 , 5] . The establishment of NTM-PD in impaired lungs can cause the destruction of the pulmonary parenchyma [8] and eventually lead to death due to the evolution of NTM-PD [9] . Patients with NTM-PD are not all treated because current treatments are often long , expensive and not without side effects [10] . NTM-PD mortality remains high in some series ranging from 25% to 40% at five years [1 , 9 , 11] . The main factors of poor outcomes identified in mortality studies at five years corresponded to an advanced age , the existence of respiratory comorbidities , radiological cavity lesions , and some mycobacteria such as M . xenopi [11 , 12 , 13] . Given the deteriorating respiratory status of patients due to the evolution of NTM-PD despite established treatment and the relatively high mortality at five years , it seemed important to search for early factors that could predict from the first year the favorable or unfavorable evolution of NTM-PD , and thus improve prognosis . Hence , the main purpose of this study was to identify factors that contribute to the clinical , radiological and microbiological improvement at one year of a cohort of 119 patients with definite ( meeting the criteria ATS/IDSA ) or probable ( one positive sputum culture ) NTM-PD , regardless of their immune status or their respiratory history . The secondary goal of this study was to report for the first time , a clinical , radiological and microbiological description of NTM-PD in a population of Afro-Caribbean patients in the French West-Indies .
This observational study received approval from the Institutional Review Board of the French learned society for respiratory medicine ( Société de Pneumologie de Langue Française; No: 2015–003 ) . All the participants gave their written consent . The parents/guardians provided written informed consent on behalf of participants below 18 years of age . This study was carried out in accordance with the principles of the Helsinki Declaration . This study was a retrospective , multicentric , observational study over a 11-year period between 2002 and 2012 in three French university hospitals ( CHU ) , two of which are located in the French West-Indies ( University Hospital of Fort de France , Martinique; and University Hospital of Pointe-à-Pitre , Guadeloupe ) , and the 3rd in Metropolitan France ( University Hospital of Montpellier , France ) . From the computerized databases of the bacteriological laboratories of these three institutions , we searched all patients over 13 years old with at least one positive culture for NTM between 2002 and 2012 . A total of 119 patients were therefore finally retained for this study regardless of their immune status . The exclusion criteria were an age below 13 years and the absence of patient consent . This was a composite endpoint defined by the disappearance at one year of respiratory symptoms and/or initial symptoms , regression or normalization at one year of the initial radiological lesions , and negative bacteriological cultures at one year . Negative bacteriological cultures were defined as at least three consecutive negative respiratory culture specimens at the end of one year . Patients were classified as having an improved status at one year only if all the three criteria were met ( vs . unimproved status if this was not the case ) . Statistical analyses were designed to determine the parameters related to the primary endpoint , i . e . , an improved status at one year . Univariate analysis was first conducted to study the independent variables related to the primary endpoint . Statistical tests used for categorical variables were the Chi-squared test or the Fisher exact test and for quantitative variables , the Student’s t-test or the Wilcoxon-Mann-Whitney test . For all statistical tests , the significance level was set at 5% and a power >90% . Independent variables with a p-value less than 0 . 2 determined by univariate analysis were retained for the multivariate model . Multivariate analysis consisted of logistic regression analysis . The dependent variable was the binary variable ( improved / unimproved status ) ; independent variables were introduced into the model using a backward regression approach . Variables with a p-value less than 0 . 05 were selected . The results were produced as odds ratios with 95% confidence intervals . The choice of multivariate logistic regression was dictated: A subgroup analysis was performed for the population infected by MAC for the HIV-positive population and by gender . Processing and statistical analysis were performed using version 3 . 3 . 2 of the R software . The libraries used in the statistical analysis with R included: base-package , stats-package , BioStatR-package , MASS-package and pwr-package .
Patients with an improved status represented 38 . 6% ( n = 46 ) . The regression of clinical symptoms at one year concerned 56 . 3% ( 67/119 ) , the disappearance or regression of radiological lesions at one year concerned 38 . 6% ( 46/119 ) of patients and negative bacteriological cultures at one year were obtained for 51 . 2% ( 61/119 ) . A statistically significant difference was revealed between the two groups for age ( p<0 . 05 ) , place of residence ( p<0 . 01 ) and the percentage of patients with HIV-positive serology ( p<0 . 02 ) . No difference was found between the two groups for the ATS/IDSA diagnostic criteria ( 58 . 6% vs . 56 . 1% , p = 0 . 93 ) . A statistically significant difference was found between the two groups ( improved / unimproved status ) in the circumstances of the disease discovery ( p<0 . 006 ) . There was no statistically significant difference between the two groups in terms of initial respiratory symptoms and initial radiological lesions . In Guadeloupe , the main NTM encountered in decreasing order were MAC , M . simiae and M . fortuitum , in Martinique , M . fortuitum followed by MAC , then M . gordonae; and in Montpellier , MAC then M . abscessus complex , followed by M . xenopi . No statistically significant difference was found between the improved / unimproved status groups for the mycobacterial species . The ATS/IDSA criteria were met for 62% of patients with MAC , 82% with M . abscessus , 50% with M . fortuitum and 45% with M . simiae . For bacteriological samples , 76% met the ATS/IDSA microbiological criteria . There was no statistically significant difference between the two groups for the ATS/IDSA microbiological criteria . The positive predictive value ( PPV ) of the ATS/IDSA microbiological criteria for definite NTM-PD was 89% ( 68/76 ) CI 95% ( 83%-94% ) . Lastly , patients who did not meet the ATS microbiological had a four-fold increased risk of death at one year ( OR = 4 . 01 , 95% CI; 1 . 40–14 . 51 , p<0 . 01 ) . No statistically significant difference was found between the two groups for treated patients , as well as in the total duration of treatment . There was a statistically significant difference in the duration of ethambutol treatment between the two groups ( p<0 . 001 , effect size: 0 . 81 , power: 0 . 99 ) . Side effects related to treatment concerned 10 of 63 patients ( 15 . 8% ) , five had minor side effects ( digestive disorders ) and five had major side effects ( three cases of drug-induced hepatitis , one case of eye damage and a kidney failure ) . These 5 patients with major side effects had to stop their therapy . No patient in our cohort benefited from associated surgical treatment . There was a statistically significant association between the absence of negative cultures and mortality at one year ( p<0 . 001 ) . The conversion rate of bacterial cultures was 60% ( 33/55 ) for MAC , 35% ( 6/17 ) for M . abscessus complex , 37% ( 6/16 ) for M . fortuitum and 72% ( 8/11 ) for M . simiae . The total number of mortalities at one year was 14 . 2% ( n = 17 ) , all belonging to the unimproved group . The average age of deceased patients ( 13 men and four women ) was 60 years±12 . 7 . We recorded 52% tobacco smokers , and 44% COPD , 29% HIV-positive and 5% cystic fibrosis patients . NTM of deceased patients were MAC ( 9/17; 52 . 9% ) , M . abscessus complex ( 4/17; 23 . 5% ) , M . kansasii ( 2/5; 40% ) and M . fortuitum ( 2/16; 12 . 5% ) . Eight patients died of unfavorable NTM-PD evolution , one patient from pulmonary embolism and two patients from COPD exacerbations . An association between the mortality rate and mycobacterial species in the study ( p = 0 . 86 ) was not found . Factors associated with an improvement at one year were the male gender ( OR = 2 . 34 ) , HIV-positive serology ( OR = 3 . 23 ) and duration of ethambutol treatment ( OR = 2 . 24 ) . For the population meeting the ATS/IDSA diagnostic criteria , the factor associated with improvement was the duration of ethambutol treatment ( OR = 2 . 45 ) . For the group of patients infected by MAC , improvement factors were associated with age under 50 years ( OR = 1 . 88 ) and duration of macrolide treatment ( OR = 3 . 27 ) . For the group of non HIV-positive patients , improvement factors were associated with Male ( OR = 3 . 54 ) and duration of ethambutol treatment ( OR = 1 . 90 ) . A total of 29 patients were included . The CD4 count was below 200 for 23 patients ( 79% ) . The discovery of NTM revealed an HIV-positive status for 98% of patients . The most frequently found species was MAC ( 58% ) . A statistically significant difference between the two groups ( improved / unimproved status ) was found for age ( p<0 . 04 ) , percentage of treated patients ( p<0 . 04 ) , negative bacteriological cultures at one year ( p<0 . 001 ) , and percentage of deaths at one year ( p<0 . 04 ) . Sixteen of the 29 HIV–positive patients ( 55% ) were treated . The percentage of patients improved on treatment was 81% ( 13/16 ) . A statistically significant difference was found between men and women for age ( p<0 . 04 ) , chronic respiratory diseases ( cystic fibrosis , COPD and bronchiectasis; p<0 . 004 ) , the incidence of sputum for women ( p< 0 . 02 ) , the type of radiological lesions and improved status at one year ( p<004 ) in favor of men .
In this retrospective multicenter study , improvement at one year in patients with definite or probable NTM-PD was associated with the duration of ethambutol treatment , HIV-positive status and male gender . For the group of patients infected with MAC , improvement was associated with the duration of macrolide treatment and an age <50 years . Identifying predictors of improvement at one year of NTM-PD is expected to optimize the management of the disease in its early stages . | Early predictive factors for a favorable development of nontuberculous mycobacterial pulmonary disease ( NTM-PD ) are important to improve management due to the high mortality of this infection at 5 years . The purpose of this study was to search for early factors that could predict at the first year , the favorable or unfavorable evolution of NTM-PD . This multicenter and retrospective study shows the importance of the duration of use of certain antibiotics ( e . g . ethambutol and macrolides ) in combination with other drugs in the one-year improvement of patients with NTM-PD . It also confirms the favorable prognosis at one year of NTM-PD patients with HIV-positive status . Identifying predictors of improvement at one year of NTM-PD is expected to optimize prognosis of the disease in its early stages . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"respiratory",
"infections",
"pulmonology",
"signs",
"and",
"symptoms",
"fungal",
"evolution",
"bacteria",
"clinical",
"microbiology",
"sputum",
"mycology",
"mucus",
"clinical",
"pathology",
"lesions",
"actinobacteria",
"prognosis",
"nontuberculous",
"mycobacteria",
"diagnostic",
"medicine",
"anatomy",
"mycobacterium",
"avium",
"complex",
"physiology",
"biology",
"and",
"life",
"sciences",
"organisms"
] | 2017 | Predictive factors for a one-year improvement in nontuberculous mycobacterial pulmonary disease: An 11-year retrospective and multicenter study |
The E3 ubiquitin ligase COP1 ( Constitutive Photomorphogenesis 1 ) is a well known component of the light-mediated plant development that acts as a repressor of photomorphogenesis . Here we show that COP1 positively regulates defense against turnip crinkle virus ( TCV ) and avrRPM1 bacteria by contributing to stability of resistance ( R ) protein HRT and RPM1 , respectively . HRT and RPM1 levels and thereby pathogen resistance is significantly reduced in the cop1 mutant background . Notably , the levels of at least two double-stranded RNA binding ( DRB ) proteins DRB1 and DRB4 are reduced in the cop1 mutant background suggesting that COP1 affects HRT stability via its effect on the DRB proteins . Indeed , a mutation in either drb1 or drb4 resulted in degradation of HRT . In contrast to COP1 , a multi-subunit E3 ligase encoded by anaphase-promoting complex ( APC ) 10 negatively regulates DRB4 and TCV resistance but had no effect on DRB1 levels . We propose that COP1-mediated positive regulation of HRT is dependent on a balance between COP1 and negative regulators that target DRB1 and DRB4 .
Resistance ( R ) protein-mediated immunity is induced when a strain-specific avirulence ( avr ) protein from the pathogen associates with a cognate plant R protein [1] . Induction of R-mediated responses is often accompanied by the formation of a hypersensitive response ( HR ) , a form of programmed cell death resulting in necrotic lesions at the site of pathogen entry [2] . HR is one of the first visible manifestations of pathogen-induced host defenses and is thought to help prevent pathogen multiplication and spread . Plants lacking cognate R proteins can activate the less robust basal defense response , also known as pathogen-associated molecular patterns ( PAMPs ) -triggered immunity ( PTI ) . In case of bacterial and fungal pathogens , PTI involves recognition of PAMPS by the host encoded pattern recognition receptors . The basal defense response against viral pathogens involves activation of the host RNA silencing pathway , which prevents viral replication and targets viral RNA for degradation ( reviewed in [3–5] . Viruses have evolved to express suppressors that target host RNA silencing components and thereby ensure replication in the host [3–5] . Interestingly , in many cases these suppressors of RNA silencing also act as avr factors and their interaction with the host R proteins leads to activation of defense responses . For example , the Arabidopsis R protein HRT [Hypersensitive response ( HR ) to Turnip crinkle virus ( TCV ) is activated by TCV coat protein ( CP ) [6 , 7] , which is a potent suppressor of the host RNA silencing pathway [8 , 9] . However , the activation of HRT does not require silencing suppressor function since CP mutants with impaired RNA silencing suppressor activity can elicit normal HR to TCV [10] . Conversely , CP mutant R8A is a functional RNA silencing suppressor that is unable to induce normal HR on Di-17 plants and therefore is virulent on Di-17 . Although RNA silencing suppressor and avr activities of CP function independent of each other , the host RNA silencing components are intricately involved in HRT-mediated resistance signaling . This includes double stranded RNA binding protein ( DRB ) 4 , which is required for the post-translational stability of HRT and thereby HRT-mediated HR and resistance to TCV . The loss-of-function mutant drb4 supports increased replication of TCV on the inoculated leaves of HRT containing plants and systemic spread to uninoculated parts [10] . Notably , HRT drb4 or other HRT containing susceptible genotypes ( like HRT sid2 , HRT eds1 ) do not accumulate viral specific small RNAs regardless of TCV levels in their inoculated leaves [10] . This suggests that R-mediated signaling recruits components of the RNA silencing pathway but does not activate the RNA silencing pathway to target viral RNA . HRT is a coiled coil ( CC ) - nucleotide binding site ( NBS ) - leucine rich repeat ( LRR ) type R protein that is activated in the presence of CP [6 , 7 , 11 , 12] , although a direct interaction between HRT and TCV CP has not been demonstrated [13] . While HRT is sufficient for HR formation , resistance to TCV is dependent on HRT and a recessive allele at a second locus , designated rrt ( regulates resistance to TCV ) [14] . Resistance to TCV is also dependent on the SA pathway [11 , 14 , 15] . Among various components of the SA pathway that regulate HRT-mediated resistance to TCV , enhanced disease susceptibility ( EDS ) 1 , which interacts with HRT , is required for potentiation of CP-triggered HR [13] . HRT is one of the few CC-NBS-LRR proteins that has a direct dependence on EDS1 . HRT also interacts with CRT1 ( Compromised for Recognition of TCV; [16 , 17] ) but unlike EDS1 , CRT1 is not associated with activation of HR [10] . Interestingly , HRT-DRB4 complex , but not HRT-EDS1 or HRT-CRT1 dissociates in the presence of CP [10] , and might play a role in activation of HRT . Besides DRB4 , the Arabidopsis genome encodes four other DRB proteins which have been characterized for their roles in RNA biology . Among these DRB1 and DRB4 facilitate DCL1- and DCL4-mediated synthesis of miRNA and trans-acting siRNAs ( tasiRNAs ) , respectively [18 , 19] . DRB2 is also involved in the biogenesis of specific miRNA subsets [20] and DRB3 and DRB5 are thought to function in the same non-canonical miRNA pathway as DRB2 [20] . HRT-mediated resistance signaling is also dependent on blue-light photoreceptors [15 , 21] and of these knocking out CRY2 and PHOT2 results in degradation of HRT . Likewise , blue-light mediated degradation of CRY2 is also associated with the degradation of HRT [15 , 21] . HRT does not interact with CRY2 but it does interact with CRY2- and PHOT2-interacting protein COP1 ( Constitutive Photomorphogenic 1 ) [15] . COP1 is an E3 ubiquitin ligase which negatively regulates photomorphogenesis [22] . Furthermore , HRT was degraded in a 26S proteasome-specific manner and pretreatment with MG132 inhibited degradation of HRT [15] . Together , these results suggested that COP1 could be responsible for degradation of HRT . Here , we examined the role of COP1 in HRT-mediated resistance signaling . Surprisingly , we find that COP1 positively regulates HRT levels , and those of DRB1 and DRB4 . Consistent with these results both DRB1 and DRB4 are required for resistance against TCV . In contrast to COP1 , a multi-subunit E3 ligase encoded by anaphase-promoting complex ( APC ) 10 negatively regulates DRB4 and TCV resistance but had no effect on DRB1 levels . Our results suggest that COP1-mediated positive regulation of HRT is dependent on a balance between COP1 and negative regulators that target DRB1 and DRB4 .
Interaction between HRT and COP1 , together with 26S proteasome-mediated degradation of HRT , suggested that COP1 might be responsible for degradation of HRT . To test this hypothesis , we crossed cop1-6 plants ( Col-0 background , susceptible to TCV , contains recessive allele of the R gene hrt ) with Di-17 ( TCV resistant ecotype , contains the R gene HRT ) . Similar to Di-17 plants , the F2 progeny from the Di-17 x cop1 cross containing at least one copy of HRT and wild-type allele of COP1 developed visible and microscopic HR following TCV infection ( Fig 1A and 1B ) . Interestingly , in contrast , all HRT/- cop1/cop1 F2 progeny showed absence of visible or microscopic HR lesions ( Fig 1A and 1B ) , which correlated with their reduced PR-1 expression ( Fig 1C ) . Furthermore , compared to HRT COP1 plants , the HRT cop1 plants supported increased replication of TCV ( Fig 1D ) . Analysis of HRT levels revealed significantly reduced HRT protein in HRT-Flag cop1 plants as compared to HRT-Flag COP1 plants ( Fig 1E ) , even though HRT transcript levels in HRT cop1 plants were comparable or higher compared to those in HRT COP1 plants ( Fig 1F ) . This suggested that lack of COP1 affected HRT protein stability . Next , we evaluated the segregation of resistant plants in Di-17 x cop1 F2 population . All hrt/hrt and ~75% of HRT/- ( homo/heterozygous for HRT ) of F2 progeny from a Di-17 x Col-0 cross showed typical crinkled leaves and drooping bolt phenotypes associated with susceptible plants . Only 25% ( homo/heterozygous for HRT , but homozygous for rrt ) of these HR-developing progeny were able to resist TCV infection and did not allow the virus to spread into uninoculated tissues . In contrast , all HRT cop1 progeny showed susceptible phenotype suggesting that COP1 positively regulated HRT-mediated resistance to TCV ( Fig 1G and 1H and S2 Table ) . Likewise , COP1 was also required for basal resistance to TCV; in comparison to Col-0 , the cop1 plants accumulated more TCV CP in their inoculated leaves ( Fig 1I ) . To determine if COP1 regulates levels of other R proteins , we analyzed the role of COP1 in RPM1-mediated resistance . The R protein RPM1 confers resistance to the avrRpm1 expressing strain of Pseudomonas syringae pv . tomato ( Pst ) [23] . To this end , we crossed cop1 with Col-0 plants expressing RPM1-Myc under its native promoter and generated cop1 RPM1-Myc plants . Interestingly , RPM-Myc protein levels were significantly reduced in the cop1 mutant background ( Fig 2A ) , even though RPM1 transcript levels in the cop1 background were comparable to those in COP1 plants ( S1A Fig ) . This suggested that lack of COP1 affected RPM1 protein stability . Consistent with this phenotype , the cop1 plants showed increased susceptibility to avrRpm1 Pst ( Fig 2B ) . Next , we assayed the interaction between COP1 and RPM1 using bi-molecular fluorescence complementation ( BiFC ) assay in Nicotiana benthamiana and co-immunoprecipitation ( IP ) assays in N . benthamiana and Arabidopsis . RPM1 did interact with COP1 and this interaction was primarily observed in the cytoplasm ( S1B Fig ) . The BiFC result was verified using IP assays of transiently expressed proteins in N . benthamiana ( Fig 2C ) and confirmed in the native Arabidopsis system ( Fig 2D ) . Together , these results suggested that COP1 positively regulates RPM1 levels . Recent results showing that COP1 and DRB4 positively regulate DRB1 [24] and HRT levels [10] , respectively , prompted us to analyze the relationship between COP1 , DRB1 and DRB4 . Interestingly , cop1 plants contained reduced levels of both DRB1 and DRB4 ( Fig 3A and 3B ) . However , the cop1 plants accumulated normal levels of DRB2 ( S1C Fig ) . The evaluation of DRB3 and DRB5 levels was limited by the inability of DRB3 and DRB5 specific antibodies to detect distinct bands in protein gel-blot analyses . A loss-of-function mutations in two proteins ( SPA1 or PIF1 ) that interact with COP1 and contribute to COP1 activity , did not affect DRB4 levels ( Fig 3B ) . This suggests that COP1 protein that was not in the COP1-SPA1-PIF1 complex contributed to DRB4 protein levels . To determine if this regulation of DRB4 levels involved interactions between COP1 and DRB4 , we generated Arabidopsis plants coexpressing COP1-Flag and DRB4-Myc and used these for IP assays . No interaction was detected between COP1 and DRB4 ( Fig 3C ) . Thus , unlike DRB1 [24] , COP1-mediated regulation of DRB4 is unlikely to be the result of direct/indirect physical associations between these proteins . Furthermore , the drb1 and drb4 mutants contained wild-type-like levels of the reciprocal proteins ( Fig 3A and 3B ) and DRB1 did not associate with DRB4 ( Fig 3D ) . Together , these results suggested that COP1-mediated regulation of DRB1 and DRB4 levels likely involves independent processes . Earlier we showed that DRB4 is required for HRT-mediated resistance to TCV signaling [10] . To test if DRB1 , and other DRB proteins , are also required for HR and/or resistance to TCV , we generated homozygous mutant lines in all DRB proteins . All the knock-out ( KO ) lines used here were characterized in a previous study [25] ( S1 Table ) . As shown before , drb1 plants showed short-stature and drb4 plants showed the zippy ( narrow leaves ) phenotype ( S2A Fig ) . Next , we crossed drb plants ( Col-0 background ) with Di-17 ( TCV resistant ecotype ) . The F2 progeny from a Di-17 x Col-0 control cross or HRT introgressed into Col-0 background ( backcrossed 8 times ) were used as controls and both these genotypes developed visible and microscopic HR following TCV infection ( Fig 3E ) . Likewise , all HRT/- drb/drb F2 progeny , except HRT/- drb1/drb1 , developed normal HR ( Fig 3E and 3F ) and induced wild-type-like PR-1 gene expression ( Fig 3G ) . In contrast , HRT/- drb1 plants showed fewer microscopic HR lesions ( Fig 3F ) , which correlated with their reduced PR-1 expression ( Fig 3G ) . Notably , all HRT drb genotypes supported increased replication of TCV compared to Di-17 or Col-0-HRT plants ( Fig 3H ) . Together , these results suggested that while all DRBs were required for HRT-mediated local resistance to TCV , only DRB1 contributed to HR development in response to TCV infection . Next , we evaluated the segregation of resistant plants in Di-17 x drb F2 population . All hrt/hrt and ~75% of HRT/- ( homo/heterozygous for HRT ) of F2 progeny from a Di-17 x Col-0 cross showed typical crinkled leaf and drooping bolt phenotypes associated with TCV susceptibility . Only 25% ( homo/heterozygous for HRT , but homozygous for rrt ) of these HR-developing progeny were able to resist TCV infection and did not exhibit virus spread to uninoculated tissues . Evaluation of genetic segregation in Di-17 x drb2 and Di-17 x drb3 crosses showed statistically significant deviation from Mendelian segregation; all HRT drb2 and HRT drb3 plants showed typical susceptible symptoms suggesting that DRB2 and DRB3 proteins were required for resistance to TCV ( S2B Fig , S2 Table ) . The involvement of DRB5 in TCV resistance could not be fully ascertained since DRB5 is located 1 Mb North of HRT resulting in skewed segregation in the progeny of Di-17 x drb5 cross ( S2 Table ) . Nonetheless , all HRT drb5 progeny showed susceptible phenotype suggesting that DRB5 was also required for HRT-mediated resistance to TCV ( S2B Fig , S2 Table ) . Likewise , involvement of DRB1 in the resistance response could not be firmly established since HRT drb1 F2 plants were difficult to inoculate due to their curled leaves and often yielded ~5–12% resistant plants ( S2 Table ) . The requirement of DRB1 in HRT-mediated resistance was further assessed using DRB1 knock-down plants ( see below ) . Earlier we showed that degradation of HRT was associated with a spreading HR phenotype wherein HR lesions coalesced resulting in prominent chlorosis [10] . This was seen in HRT drb4 , HRT crt1 , and HRT cry2 genotypes , all of which showed reduced levels of HRT [10] . Comparison of HR phenotypes in the HRT drb genotypes at 10 days post inoculation ( dpi ) showed pronounced chlorosis on HRT drb2 , HRT drb3 and HRT drb5 but not HRT drb1 leaves ( Fig 4A ) . Analysis of HRT levels revealed significantly reduced HRT protein in HRT-Flag drb1 , HRT-Flag drb2 , HRT-Flag drb3 , and HRT-Flag drb5 transgenic plants as compared to HRT-Flag DRB plants ( Figs 4B and S3A ) , even though HRT-Flag transcript levels in HRT-Flag drb plants were comparable to those in wild-type plants ( S3B Fig ) . This suggests that absence of DRB proteins specifically affected HRT protein stability . Together with the spreading HR phenotype of HRT drb plants , this suggests that a certain threshold level of HRT is required for proper HR . Clearly , the spreading HR phenotype was HRT-dependent and unrelated to TCV replication because Col-0 plants ( hrt ) , which contained the highest levels of TCV in inoculated leaves , did not show spreading lesions/cell death ( Figs 3E and 4A ) . Likewise , hrt drb plants did not show HR lesions , and Col-0-HRT plants showed Di-17-like non-spreading HR-like lesions ( S3C Fig ) . To determine if DRB proteins contribute to the stability of HRT via physical interactions with the R protein , we used BiFC assays in N . benthamiana . HRT did interact with DRB1 , DRB3 , and DRB5 , but not DRB2 , and these interactions were primarily observed in the cytoplasm ( S3D Fig ) . The BiFC results were verified using IP assays of transiently expressed proteins in N . benthamiana ( S3E , S3F , S3G and S3H Fig ) . The IP assays for DRB1 and DRB2 were further confirmed in the native Arabidopsis system where DRB1 and DRB2 proteins were expressed under their native promoters ( Fig 4C and 4D and S1 Table ) . To determine if increased expression of DRB proteins potentiated the activation of HRT , we first monitored the HR phenotype in N . benthamiana plants transiently co-expressing DRB1 , DRB3 , or DRB4 with HRT and CP ( Fig 4E ) . As shown earlier , co-expression of HRT and CP triggered nominal cell death , and the presence of EDS1 enhanced this response [13] ( Fig 4E ) . Interestingly , co-expression of DRB1 , DRB3 or DRB4 proteins with HRT and CP also enhanced HR ( Figs 4E and S3I ) . To confirm this in the native system we generated Arabidopsis plants overexpressing DRB1 , DRB3 , or DRB4 , in the Di-17 background and evaluated T2 and T3 plants for HR and resistance to TCV . Multiple lines were evaluated for each transgene and at least two lines expressing higher levels of DRB transcripts were selected for further analysis ( S3J Fig ) . As observed in transient assays , overexpression of DRB1 , DRB3 or DRB4 resulted in increased cell death response after TCV infection and this phenotype was particularly pronounced in DRB1 overexpressing plants ( Fig 4F and 4G ) . Notably , this analysis also identified two Di-17 DRB1 lines that showed significantly reduced expression of DRB1 ( #1–1 and 1–8 , S3J Fig ) , likely due to transgene co-suppression . Interestingly , like HRT drb1 , the DRB1-1 and DRB1-8 lines showed impaired HR ( S3K Fig ) , which corresponded to increased susceptibility to TCV ( S3L , S3M and S3N Fig ) . All 35S-DRB plants showed wild-type-like susceptibility to the virulent TCV strain R8A ( S3O–S3Q Fig ) . Together , these results suggested that DRB proteins are important for stabilizing HRT and that DRB1 plays a more important role in HRT-mediated signaling . This is further consistent with the impaired activation of HRT in cop1 plants , which contains reduced levels of DRB1 protein . Since HRT interacts with DRB1 , DRB3 , DRB4 , DRB5 and COP1 , it was possible that degradation of HRT in drb plants is due to impaired COP1 function . To test this we evaluated photomorphogenesis in drb plants . As expected , cop1 plants were unable to sense light and produced a short hypocotyl when grown in the dark ( S4A and S4B Fig ) . In comparison , wild-type and drb mutant plants produced a long hypocotyl in the dark suggesting that drb plants are not impaired in the COP1 function ( S4A and S4B Fig ) . Unlike DRB2 , DRB3 , or DRB5 , the DRB1 protein preferentially localizes to the nucleus in transient assays carried out in N . benthamiana ( S5A Fig ) . However , DRB1 interacts with HRT in the cytosol ( S3D Fig ) . To follow up on this observation , we assayed the effect of TCV infection on the sub-cellular localization of DRB1 . Because we were unable to obtain native promoter-based DRB1-GFP transgenic plants , we assayed localization of DRB1 in transgenic drb1 plants expressing DRB1-Myc via its native promoter . Surprisingly , unlike our transient localization assays ( S5A Fig ) and transient assays reported by others [26–28] , a significant proportion of DRB1 was detected in the cytosol of Arabidopsis plants ( Fig 5A ) . Notably , the nuclear-cytoplasmic DRB1 levels seen in our study are consistent with an earlier report that evaluated DRB1 levels in Arabidopsis plants [24] . Interestingly , TCV infected plants showed a ~2 . 34-fold reduction in the nuclear levels of DRB1 ( normalized based on H3 levels; Fig 5A ) , suggesting cytoplasmic relocalization of some nuclear DRB1 in response to TCV infection . Co-expression of CP-RFP with DRB1-GFP in N . benthamiana also increased the extranuclear localization of DRB1 , directing DRB1 and CP to punctate foci in the cytoplasm ( shown by arrowheads , S5B Fig ) . In contrast , CP did not appear to alter the overall nuclear or extra-nuclear levels of DRB2-GFP ( S5B Fig ) . Notably , a small percentage of CP was detected in the nuclear fraction ( Fig 5A ) . This promoted us to assay the interaction between CP and DRB proteins . CP interacted with all DRB proteins in IP assays carried out in Arabidopsis and N . benthamiana Figs ( 5B , 5C and 5D and S5C , S5D , S5E and S5F ) . CP also interacted with DRB4 in the yeast-two hybrid assay ( S5G and S5H Fig ) , suggesting that CP directly associated with DRB4 . BiFC assays showed that the interaction between CP and DRB proteins was preferentially observed in inclusion structures that are formed in cells containing CP ( S5I Fig ) . To confirm that nuclear DRB does not associate with CP we assayed the interaction between CP and DRB2 that was directed exclusively to the nucleus ( fused with nuclear localization signal , NLS ) or the cytosol ( fused with nuclear export signal , NES ) . CP interacted with DRB2-NES ( cytosolic DRB2 ) but not DRB2-NLS ( nuclear DRB2 ) ( S5I and S5J Fig ) , suggesting that the CP-DRB complex occurred only in the cytosol . Earlier we showed that TCV infection ( in Arabidopsis ) , or CP expression ( in N . benthamiana ) , increased the cytosolic pool of DRB4 and inhibited the HRT-DRB4 interaction [10] . As shown above CP also increased the cytosolic levels of DRB1 ( Figs 5A and S5B ) . Therefore , we assayed HRT-DRB1 complex formation in the presence or absence of CP . Interestingly , like HRT-DRB4 , TCV infection or presence of CP also inhibited the HRT-DRB1 interaction ( Fig 5E and 5F ) . Notably , this was not the case for the HRT-DRB3 interaction ( S5K Fig ) , suggesting that CP-dependent inhibition of HRT-DRB1/DRB4 interactions was not a generalized effect . The CP-dependent dissociation of HRT-DRB1 complex correlated with impaired HR phenotype in HRT drb1 plants . However , HRT drb4 showed normal HR at 3 dpi even though CP also inhibited the HRT-DRB4 interaction [10] . This , together with the impaired HR phenotypes of HRT drb1 and DRB1 knock-down plants , suggested that DRB1 may be a dominant player in HR formation and thereby activation of HRT . Congruent with this notion , HRT cop1 plants that lack both DRB1 and DRB4 proteins showed loss of both visible as well as microscopic HR ( Fig 1B ) , suggesting that DRB1 and DRB4 acted additively , with DRB1 played a major role in the activation of HRT . To determine if degradation of DRB4 in cop1 plants occurred in a 26S proteasome-dependent manner , we assayed recovery of DRB4 in the cop1 plants that were treated with proteasome-specific inhibitor MG132 . The cop1 leaves pretreated with MG132 accumulated significantly higher levels of DRB4 protein ( Fig 6A ) , suggesting that DRB4 in cop1 plants was degraded in a proteasome-dependent manner . This is further consistent with earlier results showing that APC10 subunit of the anaphase promoting complex ( APC ) interacted with DRB4 and elevated levels of DRB4 in APC10 RNAi plants suggested that APC10 targeted DRB4 for degradation [29 , 30] . To test this further and to investigate relationship between APC10 and COP1 , we first analyzed DRB4 levels in APC10 overexpressing Col-0 plants [31] . As predicted , the APC10 overexpressing plants showed reduced levels of DRB4 , suggesting that the increased expression of APC10 negatively regulated accumulation of DRB4 ( Fig 6B , upper panel ) . Consistent with this result , the APC10 overexpressing plants showed increased levels of TCV-CP in their inoculated leaves ( Fig 6B , middle panel ) . Normal DRB4 transcript in APC10 overexpressing plants suggested that APC10-mediated negative regulation of DRB4 was a post-translational event ( Fig 6C ) . Overexpression of APC10 had no effect on DRB1 ( Fig 6B , bottom panel ) . No interaction was detected between APC10 and COP1 ( S6A Fig ) , suggesting that APC10-mediated negative regulation of DRB4 did not involve physical sequestration of COP1 . Furthermore , APC10 overexpressing plants showed wild-type like photomorphogenetic phenotype in light and dark , suggesting that these plants were not altered in the COP1 function ( S6B Fig ) . Analysis of Arabidopsis interactome comprising of predicted or known interactions with COP1 , DRB4 , and APC10 was unable to identify any proteins that are shared between COP1 and DRB4 or APC10 ( S6C Fig ) . Together , these results suggest that COP1- and APC10-mediated regulation of DRB4 might involve independent processes but the relative levels APC10 play an important role in the stability of DRB4 ( Fig 6D ) , and thereby disease resistance .
The earth’s natural light environment undergoes continuous spatial and temporal fluctuations and living organisms have evolved to regulate their growth and well-being in response to these fluctuations . Plants being sessile have to particularly modify their growth and development for optimized utilization of ambient light . Thus , it is conceivable that photobiology is integral to plant defense . Although several studies show an important role for light in plant defense [32] , the precise molecular mechanisms underlying interactions between plant immunity and light perception are less understood . Here we show that COP1 , an important master regulator that negatively regulates photomorphogenesis by degrading key proteins involved in light-regulated plant development , plays an equally important role in HRT-mediated defense against TCV and RPM1-mediated resistance against avrRPM1 bacteria . However , in contrast to its role in photomorphogenesis , COP1 functions as a positive regulator in plant defense and is required for the stability of the R proteins HRT and RPM1 . COP1 conferred regulation of HRT involves at least two DRB proteins , DRB1 and DRB4 , which are well known components of the RNA silencing machinery . A mutation in COP1 results in the degradation of DRB1 and DRB4 , which confer stability to HRT . Likewise , DRB4 is also required for the stability of RPM1 [10] , and consistent with this result RPM1 accumulates to very low levels in the cop1 plants . Intriguingly , besides DRB1 and DRB4 , three other DRB proteins also participate in TCV resistance by regulating levels of the R protein HRT . This corresponds to the physical interaction of HRT with DRB1 , DRB3 , DRB4 , and DRB5 . Normal levels of DRB2 in cop1 plants suggests that COP1 might specifically regulate DRB1 and DRB4 proteins . We were unable to determine DRB3 and DRB5 levels in cop1 plants due to lack of specific antibodies . The severely stunted phenotype of cop1 in comparison to drb1 drb4 double mutants suggests that additional components might contribute to the growth phenotype of cop1 mutant plants ( S6D Fig ) . Normal photomorphogenic response displayed by the drb mutants suggests that the loss of DRB proteins does not alter COP1 function . Thus , COP1 likely protects DRB1 and DRB4 from one or more negative regulators that target these proteins for degradation ( Fig 6D ) . Consequently , loss of COP1 would render these negative regulators active resulting in the degradation of DRB1 and DRB4 . COP1 was recently shown to stabilize DRB1 by negatively regulating an unknown protease [24] . However , unlike DRB1 , COP1 does not interact with DRB4 or APC10 E3 ligase , which negatively regulates DRB4 . Thus , it is possible that COP1- and APC10-mediated regulation of DRB4 involves independent or indirect processes . Clearly , COP1 is epistatic to APC10 in relation to their effect on DRB4 since increased levels of APC10 was able to overcome COP1-mediated positive regulation of DRB4 . Normal photomorphogenic response displayed by the APC10 overexpressing plants suggests that overexpression of APC10 does not alter COP1 function . Interestingly , no physical interaction was observed between HRT and DRB2 , even though the drb2 mutant contains little or no HRT protein similar to the other drb mutants . This presents several possibilities: 1 ) DRB2 regulates HRT levels via its presence in HRT complexes comprising other DRB proteins . In fact , DRB2 and DRB4 were shown to interact with DRB1 and DRB5 in far-western assays [27] and supported by interactome analysis ( S7A Fig ) . Although these interactions cannot be detected in planta , it is possible that they exist but are undetectable under the harsh conditions used for in planta IP assays . Weak interactions between DRB2 and DRB1/5 could enable the formation of multi-protein complexes that stabilize HRT . Loss of one or more components could disrupt such complexes resulting in the degradation of HRT . 2 ) DRB2 regulates HRT levels via its presence in HRT complexes comprising other proteins ( other than DRBs; S7A Fig ) . Indeed , DRB2 is known to interact with other plant proteins including forming large molecular weight complexes and interacting with proteins involved in the regulation of chromatin functions [33] . Whether those proteins interact with HRT and/or affect its stability is not known . Similar to DRB2 , DRB4 also forms a high molecular weight ~2 MDa complex [33] ( S7B Fig ) . 3 ) DRB2 regulates HRT levels by negatively regulating a protein that degrades HRT . Notably , a high molecular weight complex comprising DRB2 contains MSI4 , which functions as a substrate adaptor for CULLIN4 ( CUL4 ) - Damaged DNA Binding Protein1 ( DDB1 ) ubiquitin E3 ligases . Furthermore , CUL4-DDB1 interacts with the COP1 complex to regulate photomorphogenesis and flowering [34 , 35] , suggesting a potential link between DRB2 and the regulation of HRT levels via E3 ubiquitin ligases . Contrary to previous studies that examined the subcellular localization of transiently expressed DRB1 in heterologous plants [26–28] , we and Cho et al . , [24] show that the bulk of transgene-expressed DRB1 is present in the cytosol of Arabidopsis plants . Notably , TCV infection relocalized ~2 . 34-fold DRB1 from the nucleus to cytosol . This is reminiscent of DRB4 , which relocalizes from the nucleus to cytosol in the presence of TCV [10] . Interestingly , the extranuclear enrichment of DRB1 and DRB4 in TCV infected plants is associated with loss of interaction with HRT . Furthermore , even though both DRB1 and DRB4 can potentiate HRT-mediated cell death to TCV , only the drb1 mutant is impaired in HR to TCV . Thus , DRB1 likely plays a key role in the activation of HRT while DRB4 has a minor role . Indeed , HRT cop1 plants did not generate visible or microscopic HR against TCV . Thus , DRB1 and DRB4 act additively with DRB1 playing a major role in defense against TCV . NBS-LRR proteins are multi-domain R proteins , which in the absence of pathogen infection remain in an inactive state . It is thought that the activated state of the R proteins involves conformational changes that exposes the N-terminal domain and thereby allows the R proteins to interact with their signaling partners [1] . For instance , activation of Rx was proposed to involve CP-mediated disruption of intramolecular interactions [36] . Similarly , R protein MLA in barley was shown to self-associate in planta in an effector-independent manner [37] . Our combined results , that HRT can self-associate [10] , form complexes with DRB1 and DRB4 that are disrupted by CP , together with reduced stability of HRT in drb1 and drb4 backgrounds , propose a new model for R protein activation in plants . According to this model , DRB1 and DRB4 proteins help to maintain HRT in a dormant and stable state . CP-triggered dissociation of HRT-DRB complexes relieves the DRB1/4-mediated repression of HRT , facilitating a conformational change that triggers activation of HRT . It is possible that DRB2 , 3 and 5 serve as decoys for CP and this further explains the inability of CP to disrupt the HRT-DRB3 interaction . Determining the precise relationships between the different DRB proteins in regulating various aspects of plant development will help better elucidate the canonical and non-canonical functions of these proteins .
Plants were grown in MTPS 144 Conviron ( Winnipeg , MB , Canada ) walk-in-chambers at 22°C , 65% relative humidity and 14 hour photoperiod . The photon flux density of the day period was 106 . 9 μmoles m-2 s-1 and was measured using a digital light meter ( Phytotronic Inc , Earth city , MO ) . Plants were grown on autoclaved Pro-Mix soil ( Premier Horticulture Inc . , PA , USA ) . Soil was fertilized once using Scotts Peter’s 20:10:20 peat lite special general fertilizer that contained 8 . 1% ammoniacal nitrogen and 11 . 9% nitrate nitrogen ( Scottspro . com ) . Plants were irrigated using deionized or tap water . Crosses were performed by emasculating the flowers of the recipient genotype and pollinating with the pollen from the donor . F2 plants showing the wt genotype at the mutant locus were used as controls in all experiments . The wt and mutant alleles were identified by PCR , CAPS , or dCAPS analysis . The Col-0-HRT line was generated after eight backcrosses of F1 derived from a Di-17 x Col-0 cross with Col-0 , which was used as a recurrent parent . The F1 and F2 progenies from each backcross were genotyped for HRT and those from initial and final backcrosses were tested for HR and resistance phenotypes . The Di-17 and Col-0 transgenic plants expressing HRT-Flag transgene are described earlier [15] . For transgenic overexpression of DRBs , the cDNA spanning the coding region were cloned into the pGWB5 vector [38] , and expressed using 35S promoter and NOS terminator . The transgenic plants were selected on plates containing kanamycin ( 50 μg/ml ) and hygromycin ( 17 μg/ml ) . For native expression of DRBs , the Myc or Flag-HA tagged DRBs along with their respective promoters were cloned into pCambia 1300 derived vector and transformed into respective drb mutant backgrounds . Genetic complementation was assayed by analyzing the levels of siRNA , as described before [33] . Small-scale extraction of RNA from two or three leaves ( per sample ) was performed with the TRIzol reagent ( Invitrogen , CA ) , following the manufacturer’s instructions . RNA gel blot analysis and synthesis of random-primed probes for PR-1 , CP and DRB4 were carried out as described previously [14] . RNA quality and concentration were determined by gel electrophoresis and determination of A260 . Reverse transcription ( RT ) and first strand cDNA synthesis were carried out using Superscript II ( Invitrogen , CA ) . Quantitative RT-PCR was carried out as described before [39] . Each sample was run in triplicates and ACTIN II ( At3g18780 ) expression levels were used as internal control for normalization . Cycle threshold values were calculated by SDS 2 . 3 software . The leaves were vacuum-infiltrated with trypan-blue stain prepared in 10 mL acidic phenol , 10 mL glycerol , and 20 mL sterile water with 10 mg of trypan blue . The samples were placed in a heated water bath ( 90°C ) for 2 min and incubated at room temperature for 2–12 h . The samples were destained using chloral hydrate ( 25 g/10 mL sterile water; Sigma ) , mounted on slides and observed for cell death with a compound microscope . The samples were photographed using an AxioCam camera ( Zeiss , Germany ) and images were analyzed using Openlab 3 . 5 . 2 ( Improvision ) software . Transcripts synthesized in vitro from a cloned cDNA of TCV using T7 RNA polymerase were used for viral infections . For inoculations , the viral transcript was suspended at a concentration of 0 . 05 μg/ μL in inoculation buffer , and the inoculation was performed as described earlier [40] . After viral inoculations , the plants were transferred to a Conviron MTR30 reach-in chamber maintained at 22°C , 65% relative humidity and 14 hour photoperiod . HR was determined visually three-to-four days post-inoculation ( dpi ) . Resistance and susceptibility was scored at 14 to 21 dpi and confirmed by northern- or western-gel blot analysis . Susceptible plants showed stunted growth , crinkling of leaves and drooping of the bolt . The bacterial strain pVSP61 ( empty vector ) , or avrRpm1 were grown overnight in King’s B medium containing rifampicin and kanamycin ( Sigma , MO ) . The bacterial cells were harvested , washed and suspended in 10 mM MgCl2 . The cells were diluted to a final density of 105 or 106 CFU/mL ( A600 ) and used for infiltration . The bacterial suspension was injected into the abaxial surface of the leaf using a needle-less syringe . Three leaf discs from the inoculated leaves were collected at 0 and 3 or 6 dpi . The leaf discs were homogenized in 10 mM MgCl2 , diluted 103 or 104 fold and plated on King’s B medium . Proteins were extracted in buffer containing 50 mM Tris-HCl , pH7 . 5 , 10% glycerol , 150 mM NaCl , 10 mM MgCl2 , 5 mM EDTA , 5 mM DTT , and 1 X protease inhibitor cocktail ( Sigma-Aldrich , St . Louis , MO ) . Protein concentration was measured by the Bio-RAD protein assay ( Bio-Rad , CA ) . For small scale extractions 2–3 leaves were homogenized per sample . For Ponceau-S staining , PVDF membranes were incubated in Ponceau-S solution ( 40% methanol ( v/v ) , 15% acetic acid ( v/v ) , 0 . 25% Ponceau-S ) . The membranes were destained using deionized water . Proteins ( 30–50 μg ) were fractionated on a 7–10% SDS-PAGE gel and subjected to immunoblot analysis using α-CP , α-Myc , α-Flag ( Sigma-Aldrich , St . Louis , MO ) or α-GFP antibody . Immunoblots were developed using ECL detection kit ( Roche ) or alkaline-phosphatase-based color detection . Fold change , normalized with Rubisco , Actin or H3 proteins , in western blots was quantified using Image Quant software . Coimmunoprecipitations were carried out as described earlier [13 , 15] . Briefly , ~1 g of infiltrated leaf tissues were harvested and extracted in buffer containing 10% ( v/v ) glycerol , 25 mM Tris-HCL pH 7 . 5 , 1 mM EDTA , 150 mM NaCl , 2% ( w/v ) polyvinylpolypyrrolidone and 1 X protease inhibitor cocktail . Extracts were centrifuged twice at 12 , 000 g for 10 min at 4°C and supernatant was incubated overnight with 20 μl of anti-Flag M2 or anti-Myc affinity beads ( Sigma-Aldrich , St . Louis , MO ) . Beads were washed three times with the extraction buffer and the proteins were fractionated on SDS-PAGE gels as described above . Nuclear fractionation was carried out as described before [41] . For gel filtration experiments , ground mixed flower tissues were dissolved in lysis buffer ( 150 mM NaCl , 0 , 1% Igepal , 50 mM Tris pH8 , 5 mM MgCl2 , 10 μM MG132 1X protease inhibitor cocktail ) . Supernatant was filtered through 0 . 45 μm membrane , and further processed by a 2 hours centrifugation , 4500 rpm on Amicon Ultra centrifugal units ( Millipore ) . 500μl of the resulting crude extract was loaded onto the Superose 6 10/200 column ( GE Healthcare ) to perform size exclusion chromatography , 500μl/minute , and 500μl fractions were collected , precipitated separately in 2 volumes of absolute Ethanol overnight at 4°C , and pellets were resuspended in 100ul 2X Laemmli buffer . Separation and blotting was then performed as described above . Size markers were run in similar settings , in a separate run . For confocal imaging , samples were scanned on an Olympus FV1000 microscope ( Olympus America , Melvile , NY ) . GFP ( YFP ) , and RFP were excited using 488 , and 543 nm laser lines , respectively . Constructs were made using pSITE [42] , pEarlyGate or pGWB based binary vectors using Gateway technology and introduced in A . tumefaciens strain LBA4404 for agroinfiltration into N . benthamiana and MP90 for Arabidopsis transformation . Agrobacterium strains carrying various constructs were infiltrated into wild-type or transgenic N . benthamiana plants expressing CFP-tagged nuclear protein H2B . 48 h later , water-mounted sections of leaf tissue were examined by confocal microscopy using a water immersion PLAPO60XWLSM 2 ( NA 1 . 0 ) objective on a FV1000 point-scanning/point-detection laser scanning confocal 3 microscope ( Olympus ) equipped with lasers spanning the spectral range of 405–633 nm . RFP , CFP and GFP/YFP overlay images ( 40X magnification ) were acquired at a scan rate of 10 ms/pixel . Images were acquired sequentially when multiple fluorophores were used . Olympus FLUOVIEW 1 . 5 was used to control the microscope , image acquisition and the export of TIFF files . | Plants must constantly regulate the allocation of resources between photomorphogenesis and defense signaling . Although light is known to influence plant defense , the underlying mechanisms remain largely unknown . Here we show that light plays specific and direct signaling roles in plant defense . Specifically , a positive role for COP1 E3 ligase , an important regulator of photomorphogenesis , in plant defense is demonstrated . We further show that COP1 regulates the levels of double-stranded RNA binding proteins DRB1 and DRB4 , which in turn regulate the levels of resistance protein HRT that confers resistance against turnip crinkle virus . In contrast to COP1 , a multi-subunit E3 ligase encoded by anaphase-promoting complex ( APC ) 10 negatively regulates DRB4 but had no effect on DRB1 levels . Together , these results suggest that COP1-mediated positive regulation of HRT is dependent on a balance between COP1 and negative regulators that target DRB1 and DRB4 . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"biotechnology",
"plant",
"anatomy",
"rna",
"interference",
"protein",
"interactions",
"brassica",
"plant",
"physiology",
"protein",
"expression",
"plant",
"science",
"model",
"organisms",
"genetically",
"modified",
"plants",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"epigenetics",
"plants",
"genetic",
"engineering",
"arabidopsis",
"thaliana",
"research",
"and",
"analysis",
"methods",
"genetically",
"modified",
"organisms",
"genetic",
"interference",
"proteins",
"gene",
"expression",
"leaves",
"molecular",
"biology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"rna",
"eukaryota",
"plant",
"and",
"algal",
"models",
"photomorphogenesis",
"nucleic",
"acids",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"plant",
"biotechnology",
"organisms"
] | 2018 | COP1, a negative regulator of photomorphogenesis, positively regulates plant disease resistance via double-stranded RNA binding proteins |
The shaping of individual cells requires a tight coordination of cell mechanics and growth . However , it is unclear how information about the mechanical state of the wall is relayed to the molecular processes building it , thereby enabling the coordination of cell wall expansion and assembly during morphogenesis . Combining theoretical and experimental approaches , we show that a mechanical feedback coordinating cell wall assembly and expansion is essential to sustain mating projection growth in budding yeast ( Saccharomyces cerevisiae ) . Our theoretical results indicate that the mechanical feedback provided by the Cell Wall Integrity pathway , with cell wall stress sensors Wsc1 and Mid2 increasingly activating membrane-localized cell wall synthases Fks1/2 upon faster cell wall expansion , stabilizes mating projection growth without affecting cell shape . Experimental perturbation of the osmotic pressure and cell wall mechanics , as well as compromising the mechanical feedback through genetic deletion of the stress sensors , leads to cellular phenotypes that support the theoretical predictions . Our results indicate that while the existence of mechanical feedback is essential to stabilize mating projection growth , the shape and size of the cell are insensitive to the feedback .
From cell division to polarization and growth , cells constantly change their shapes to perform specific tasks [1–3] . These morphological changes are achieved through remodeling of the structures that mechanically sustain the cell , such as the cytoskeleton in animal cells and the cell wall in walled cells . Unlike animal cells , which can undergo fast and complex cell shape changes , walled cells must take extra care during shape changes , as the cell wall needs to mechanically sustain their high internal turgor pressure throughout the cell wall remodeling process [4–6] . A lack of coordination between cell wall expansion and assembly during cell growth can be fatal for the cell , as the thinning of the cell wall in expanding regions may lead to cell lysis unless it is carefully balanced by newly assembled wall material . While it is believed that the coordination of cell wall expansion and assembly is necessary to cell wall remodeling and morphogenesis , the mechanisms behind this coordination remain largely unknown . Cell shape changes are ultimately governed by the mechanical state of the cell wall [5–7] . Studies of the mechanics of walled cell morphogenesis have predominantly focused on tip-growing cells of plant and fungal species because of their large size , simpler geometry and fast growth rates [8–10] . In this highly polarized growth mode , cells adopt a tubular shape that extends only at the apical region ( Fig 1 ) . During this process , cells polarize their cytoskeleton and localize exocytosis to the growing region , exactly where the cell wall needs to be assembled and remodeled . While the molecular underpinnings of tip-growth differ across species , two basic features have been shown to be necessary [7]: polarized assembly of new cell wall material at the tip , and inhomogeneous mechanical properties enabling its apical expansion ( Fig 1F ) . Previous theoretical descriptions of tip-growth focused on cell wall assembly [11–13] or cell wall mechanics [10 , 14 , 15] separately . More recent descriptions accounted for both cell wall assembly and mechanics [16–19] , but assumed these processes to be independent of each other . As we show below by directly solving the dynamics of cell wall assembly and expansion , assuming cell wall mechanics and assembly to be independent of each other always leads to unstable cell wall expansion and cell lysis , in stark contrast with experimental observations . Despite its relevance to cell viability during cell wall remodeling and morphogenesis , no previous theoretical descriptions have addressed the role of coordination ( coupling or feedback ) between cell wall mechanics and assembly in the morphogenesis of walled cells . In addition to well-known model systems for tip-growth , such as pollen tubes in plants and hyphal growth in higher fungi [7–9] , budding yeast displays tip-growth during mating . Haploid cells secrete pheromone ( α- and a-factors for mating types a and α , respectively ) that elicits the growth a tubular mating projection from the partner of the opposite type [20 , 21] ( Fig 1A ) . Since the molecular basis of cell polarization and cell wall assembly and remodeling have been extensively studied in budding yeast , tip-growth of mating projections provide a unique system to study the mechanism of coordination between cell wall mechanics and assembly . In a-cells , binding of α-factor to its cognate receptor activates the heterotrimeric G-protein , leading to the activation and polarization of the small G-protein Cdc42 , a master regulator of cell polarization [22] . Cdc42-mediated polarization recruits various molecular factors to an apical region of the plasma membrane known as the polarisome , where the formin Bni1 drives the nucleation of actin cables , focusing exocytosis at the apex [3 , 20] ( Fig 1B ) . Secretory vesicles transporting Fks1/2 cell wall synthases and cell wall remodeling enzymes ( e . g . , glucanases ) move along actin cables to the exocyst , eventually leading to the incorporation of Fks1/2 synthases to the plasma membrane and the release of glucanases into the preexisting cell wall ( Fig 1B and 1C ) [23–26] . Together , these events molecularly and mechanically polarize the cell , causing a localized expansion of the cell wall at the apex ( Fig 1B and 1F ) . In general , the expansion of the cell wall is a dangerous situation that the cell needs to carefully control . Since the cell wall sustains the high cell’s internal turgor pressure , uncontrolled cell wall expansion can lead to cell wall piercing and cell lysis . In budding yeast , the Cell Wall Integrity ( CWI ) pathway is known to help the cell prevent loss of cell wall mechanical integrity in a variety of situations [27–30] , from mating pheromone-induced growth to vegetative growth [28 , 29 , 31] . Five transmembrane proteins , namely Wsc1 , Wsc2 , Wsc3 , Mid2 , and Mtl1 , are thought to act as stress sensors and relay information about the mechanical state of the cell wall to multiple intracellular processes via the activation of Rho1 GTPases [28 , 29 , 32–37] . Previous works have shown that Wsc1 and , especially , Mid2 play an important role during mating pheromone induced growth , while the remaining stress sensors do not seem to strongly affect projection growth [37–41] . While the specific mechanical quantity that these stress sensors monitor in the cell wall remains unclear , activation of the CWI pathway leads to the downstream Rho1-mediated activation of several key molecular components , including cell wall synthases ( Fks1/2 ) , actin nucleators ( Bni1 ) and mediators of exocytosis ( Sec3 ) , and also induces a transcriptional response via a MAPK cascade [29] . The activation of cell wall Fks1/2 synthases [21 , 29 , 33] provides the most direct coupling between cell wall mechanics and assembly and could potentially stabilize mating projection growth ( Fig 1D ) . However , it is unknown if such a simple , direct mechanical feedback can stabilize morphogenesis of walled cells by itself . Using mating projection growth in budding yeast as a model system , and combining experiments and theory , we show that coordination between cell wall mechanics and assembly through direct Fks1/2 activation in the CWI pathway ( mechanical feedback ) stabilizes mating projection growth without affecting its geometry . In what follows , the term ‘mechanical feedback’ refers to the nature of the input signal that is sensed and relayed by stress sensors in the CWI pathway . We first derive a theoretical description that connects the cell wall mechanics to the intracellular processes building the wall ( Fks1/2 activation dynamics ) via the CWI pathway , and show that stable projection growth can only persist in the presence of mechanical feedback . In the absence of coordination between cell wall assembly and mechanics , cell wall expansion is always unstable , leading to either progressive thickening or thinning of the cell wall depending on conditions . Our experimental results indicate the compromising the mechanical feedback through genetic deletions of the wall stress sensors Mid2 and Wsc1 , and also through perturbations of cell wall mechanics and increased turgor pressure , all lead to defects in mating projection growth and cell viability . Our experimental observations are in agreement with the theoretical predictions , suggesting that the mechanical feedback provided by the CWI pathway via direct activation of Fks1/2 synthases can stabilize projection growth without altering cell geometry . In addition , by directly measuring the size of the exocytosis region in wild-type ( WT ) and mutants with compromised mechanical feedback , we show that the size of the mating projection is controlled by the size of the exocytosis region , but is independent of the strength of the mechanical feedback , as predicted theoretically . Altogether , our results show that a mechanical feedback between cell wall mechanics and assembly is essential for stability of cell wall expansion and projection growth , but that its geometry and size are insensitive to the mechanical feedback .
The system of Eqs 1–5 was scaled and written in a manner such that r , h , ρA , and ρI were described by equations evolving in time , and u , θ , κs by differential equations in s . The latter equations were solved by the method of lines; s was discretized and the s-derivatives were written as a differential matrix using fourth order central difference and one sided finite differences at the boundary . The resulting system becomes a differential algebraic system ( DAE ) , which was solved using Sundials , a suite of nonlinear and DAE solvers . Steady state solutions were obtained by ensuring that all time derivatives of scaled variables were below 10−3 . All yeast strains were derivatives of W303-1A and contained the bar1Δ mutation that prevents α-factor degradation by deletion of the Bar1 protease . Genetic techniques were performed per standard methods [42] . Yeast strains used in this study are listed in Table A in S1 File . All strains were cultured in YPD ( yeast extract-peptone-dextrose ) media supplemented with adenine . The wsc1Δmid2Δ strain was grown in YPD media with 1M sorbitol to increase viability . Gene deletions and GFP-tagging were constructed by genomic integration using vectors amplified and targeted by PCR primers [43] . Cell lysis was determined by propidium iodide ( Molecular Probes ) staining . Propidium iodide ( PI ) was prepared in DMSO at a concentration of 20 mM and then diluted 1:1000 for use . Propidium iodide was added to cells after being exposed to α-factor ( 1 μM ) for 2 hours . To observe the viability of cells after altering the osmotic pressure , we diluted the YPD media with distilled water upon addition of propidium iodide . The cells were imaged on slides after being exposed to propidium iodide for 10 minutes . Brightfield and fluorescent ( RFP filter set ) images were acquired using an inverted Nikon Eclipse TE300 microscope with a 60x objective ( NA = 1 . 4 ) . Image analysis was manually performed using ImageJ . Data from 3 samples for each condition was averaged and , for each sample , 150 cells or more were analyzed . To decrease the viscosity of the cell wall , we utilized zymolyase , which contains β-1 , 3 glucanase , to hydrolyze the glucan linkages that strengthen the wall . Zymolyase ( Zymo Research , 1 μl ( 2 units ) per 100 μl of cells ) was added to cells exposed to alpha-factor for 1 . 5 hours . Cells were treated additionally with concanavalin A to immobilize them during the imaging process . The cells were imaged on slides for 7 minutes after being exposed to zymolyase for 3 minutes . DIC images were acquired every 3 seconds . Data from 5 samples for each condition was averaged and , for each sample , 15 cells or more were analyzed . Image analysis was manually performed using ImageJ . The length-scale of exocytosis was measured in strains that contained Sec3 fused to GFP . Calcofluor White Stain ( Sigma-Aldrich ) was added to cells 10 minutes prior to imaging ( final concentration 0 . 1mg/ml ) to distinguish the cell wall during image analysis . To properly visualize the length-scale and reduce imaging noise , we averaged 30 consecutive confocal images , taken at 2 second time intervals , for each cell , after incubation in 1 μM α-factor for 1 hour and 30 minutes . For spa2Δ cells , the 30 images were taken at 13 second intervals to average over a longer time period to average out the stronger fluctuations in polarization in this mutant . Images were acquired with a laser-scanning confocal microscope ( Zeiss LSM 710 ) , using a 100x objective ( NA = 1 . 4 ) . The cells were immobilized to a glass-bottom dish coated with concanavalin A . To horizontally orient the mating projections , we layered a YPD ( supplemented with 1 μM α-factor ) agarose pad on top of the cells . Image analysis was manually performed using ImageJ .
The expansion of the cell wall during morphogenesis is powered by the cell’s internal turgor pressure , P . Such high pressure is mechanically sustained by the cell wall , which provides mechanical integrity to the cell at all stages , including during mating projection growth . Similarly to other organisms [5 , 7 , 9] , the cell wall in budding yeast can be considered a thin shell surrounding the cell , as the wall thickness ( ∼100 nm [44] ) is much smaller than the radius of the projection ( ∼1μm [45] ) . Since the cell’s shape is determined by the location of its cell wall , we describe the growth of the mating projection as the expansion of an axisymmetric thin shell , parametrized by the arclength s from the projection apex and azimuthal angle ϕ ( Fig 1E ) . The shape of the projection is characterized by its local radius , r ( s , t ) , and the principal curvatures κs = ∂θ/∂s and κϕ = sinθ/r , respectively , where θ ( s , t ) is the angle between the local outward normal and the axis of growth ( Fig 1E ) . The coordinates ( r , ϕ , z ) ( Fig 1E ) are standard cylindrical coordinates , and the angle θ and arclength s parameterize changes in normal and tangential directions of the surface , n ^ and s ^ respectively [19 , 46] ( Fig 1E ) . The time evolution of the mating projection shape is governed by the mechanics and assembly of the cell wall , as described below . In the absence of mechanical feedback ( Γ = 0 ) and only active Fks1/2 ( see S1 File for details ) mating projection growth is unstable for any value of the parameters . We find that in the absence of mechanical feedback the cell wall either progressively thins , eventually leading to either cell wall piercing , or thickens , leading to unbounded cell wall growth , depending on parameter values ( S1 File ) . This instability arises from the lack of coordination between cell wall expansion and assembly: changes in cell wall expansion cannot be balanced by cell wall assembly unless the processes building the cell wall have information about how cell wall expansion is changing on the cell’s surface . In the presence of mechanical feedback ( Γ > 0 ) , numerical integration of Eqs 1–5 ( Methods ) shows that stable states of mating projection growth can be sustained for a large range of parameters ( Fig 2F and S1 File ) . In this context , stable states refer to sustained steady state growth of the mating projection at constant velocity . For any given value of the ratio ( PρwλX ) / ( 12μ0mwρ0kp ) there exists a critical value of the feedback strength Γ below which mating projection growth is unstable . Similarly , for every value of the feedback strength Γ , there is a maximal value of ( PρwλX ) / ( 12μ0mwρ0kp ) above which mating projection growth becomes unstable . This instability is caused by the progressive thinning of the apical cell wall , eventually causing the piercing of the cell and leading to cell lysis . The bifurcation between stable and unstable states characterizes the transition between stably growing mating projections and a situation in which this stable growth cannot be sustained because of the progressive thinning of the cell wall and its eventual piercing . This instability threshold ( bifurcation ) is equivalent to the existence of a maximal turgor pressure ( or a minimal viscosity ) , above ( below ) which the cell wall progressively thins and eventually pierces at the tip of the projection , leading to cell lysis . The predicted increase of the maximal turgor pressure or decrease in the minimal wall viscosity for increasing feedback strength Γ indicates that cells with compromised mechanical feedback should be more sensitive to both an increase in turgor pressure or a decrease in wall viscosity than WT cells . In order to experimentally explore the predicted dynamical regimes ( Fig 2F ) , we systematically varied the mechanical feedback strength , as well as the turgor pressure P and cell wall viscosity μ0 . In contrast to previous works , here we examine all three perturbations in the context of the stability of pheromone-induced projection growth . We first varied the feedback strength Γ by compromising the ability of the cell to sense the mechanical state of the wall . To this end , we genetically deleted the two primary cell wall stress sensors present during mating projection growth , namely Wsc1 and Mid2 [33] ( Fig 1D ) , and measured the resulting cell lysis ( Methods ) . Only in the presence of α-factor and mating projection growth , did the deletion of either of the two sensors ( Mid2 , Wsc1 ) lead to increased levels of cell lysis compared to WT ( Fig 2A and 2E ) , as predicted theoretically ( Fig 2F ) , indicating that the ability to sense the mechanical state of the wall is essential during growth . Moreover , the double mutant mid2Δwsc1Δ exhibited the highest level of cell lysis in α-factor and , even when osmotically supported by 1M sorbitol , showed a substantial increase in lysis after the addition of α-factor ( Fig 2A , 2C and 2E ) . These observations show that the double mutant has an enhanced sensitivity to the addition of mating pheromone , in agreement with previous results obtained during vegetative growth [40] . To explore how changes in the parameter ( PρwλX ) / ( 12μ0mwρ0kp ) affected cell viability ( Fig 2F ) , we independently changed the turgor pressure P and the cell wall viscosity μ0 . To increase the cell’s turgor pressure P , we progressively decreased the osmolarity of the external medium ( Methods ) . We observed a monotonic increase in lysed cells for both WT and mid2Δwsc1Δ cells as media osmolarity was decreased in the presence of α-factor ( Fig 2B and 2D ) , consistent with the theoretically predicted effect of increased turgor pressure P ( Fig 2F ) . Finally , in order to decrease the cell wall viscosity μ0 , thereby increasing the value of the parameter ( PρwλX ) / ( 12μ0mwρ0kp ) ( Fig 2F ) , we added zymolyase to the culture media ( Methods ) . Zymolyase enzymatic activity degrades 1 , 3-β glucans in the cell wall , effectively lowering the cell wall viscosity . Addition of zymolyase led to the piercing of the cell wall typically at the tip of the mating projection ( Fig 2G and Supplementary Video ) , as expected theoretically ( Fig 2F ) . Since zymolyase will continuously degrade the cell wall , leading to the eventual piercing and lysis of all cells , we studied the temporal increase in pierced cells . Our results indicate that mid2Δ cells with reduced mechanical feedback pierced faster than WT cells when grown at the same zymolyase concentration ( Fig 2H ) , as theoretically expected ( Fig 2F ) . Overall , our experimental results are in agreement with our theoretical predictions ( Fig 2F ) and are consistent with the CWI pathway providing the necessary mechanical feedback to coordinate cell wall expansion and assembly . Stable , steady-state solutions for mating projection growth show that the shape of the mating projection is largely insensitive to variations in the feedback strength Γ and the ratio ( PρwλX ) / ( 12μ0mwρ0kp ) ( Fig 3A , 3B , 3E and 3F ) . The size ( radius ) R of the mating projection increases linearly with the size of the exocytosis region λX , but it is independent from the feedback strength Γ ( Fig 3B ) . Beyond projection shape and size , the cell wall expansion rate , ϵ ˙ s + ϵ ˙ ϕ , is always maximal at the projection apex ( s = 0 ) and decreases away from it ( Fig 3C ) , eventually vanishing as no wall expansion occurs far away from the growing apical region . The cell wall expansion rate at the projection apex , ( ϵ ˙ s + ϵ ˙ ϕ ) | s = 0 , increases with increasing turgor pressure ( or ( PρwλX ) / ( 12μ0mwρ0kp ) equivalently ) and with decreasing mechanical feedback strength ( Fig 3C and 3D ) . In contrast , the apical cell wall thickness displays the opposite behavior ( Fig 3G and 3H ) , decreasing for increasing P or decreasing Γ . These results indicate that cells closer to the instability threshold display stronger apical cell wall expansion rates and thinner cell wall ( Figs 2E and 3D and 3H ) , suggesting the strong cell wall expansion and thinning at the apex as the cause of the loss in cell wall mechanical stability . Regarding cell wall assembly during stable , steady-state projection growth , our theoretical results indicate maximal cell wall assembly at the expanding apical region . Both the total surface density of Fks1/2 synthases , ρA + ρI , and the surface density of only active Fks1/2 synthases , ρA , are maximal at the apex and decrease away from it until they vanish ( Fig 4A , 4B , 4E and 4F ) , as expected from the apically-localized exo- and endo-cytosis profiles . The apical value of the total ( or only active ) Fks1/2 surface density , namely ρ A 0 + ρ I 0 ( or ρ A 0 ) , can be either smaller or larger than the surface density ρ0 of Fks1/2 synthases secreted by exocytic vesicles ( Fig 4A , 4B , 4E and 4F ) . The reason why the total Fks1/2 surface density ρ A 0 + ρ I 0 can be larger than ρ0 at the apex is that active Fks1/2 is secreting 1 , 3-β glucans into the cell wall , a process that effectively anchors them to the wall , holding secreted Fks1/2 synthases to the tip region and increasing its concentration there . Beyond Fks1/2 , anchoring transmembrane proteins to the cell wall can potentially be used as a mechanism to locally increase the protein concentration on the membrane to levels well-beyond secretion levels . The fraction of active Fks1/2 , ρA/ ( ρA + ρI ) , is also maximal at the apex and decreases away from it ( Fig 4C and 4D ) . This is because of mechanical feedback , which induces more Fks1/2 activation at the apex following the larger cell wall expansion rate in this region ( Fig 3C ) . Finally , the surface concentration of inactive Fks1/2 also decreases away from the expanding tip because of tip-localized exo- and endo-cytosis ( Fig 4G and 4H ) . Non-monotonic profiles of inactive Fks1/2 occur because high cell wall expansion rates at the tip lead to more Fks1/2 activation , leaving less inactive Fks1/2 molecules in this region . Altogether , these results indicate that at the instability threshold , the apical cell wall expansion rate becomes too large to be balanced by cell wall assembly , leading to the progressive thinning of the cell wall and cell lysis . The theoretical results above predict that both the geometry and size of the growing mating projection are independent from the mechanical feedback strength Γ , and that the projection radius increases with the size of the exocytosis region ( Figs 5A and 5B and 3A and 3B ) . To experimentally explore how exocytosis and the mechanical feedback strength affect the mating projection size R ( Fig 5A ) , we employed a deletion mutant for Spa2 , a scaffold protein that localizes Bni1 and is recruited by Cdc42 [23] , which displays a very wide mating projection compared to WT ( Fig 5C and 5D ) . We visualized the exocytosis region in both WT and spa2Δ cells by expressing GFP-tagged Sec3 , a component of the exocyst that marks exocytic sites [58] . The exocytosis length scale λX ( Fig 5A ) , which we measured directly from confocal images ( Fig 5C and 5D and Methods ) , is considerably larger in spa2Δ mutant cells than in WT cells ( Fig 5E ) , indicating that a larger mating projection radius R is associated with a larger size of the exocytosis region . In contrast , the size R of the mating projection was not observed to vary with changes in the strength of the mechanical feedback Γ ( Fig 5E ) , as shown by deleting Mid2 or Wsc1 in both WT and spa2Δ backgrounds , while simultaneously measuring the size of the mating projection R and the length of the exocytosis region using Sec3-GFP . While deletion of Wsc1 and Mid2 strongly affects mating projection stability ( Fig 2E ) , our measurements show that it does not affect the size of the mating projection ( Fig 5E ) . These results indicate that the mechanical feedback is essential to sustain stable mating projection growth , but it does not affect mating projection size , which is controlled by the exocytosis profile , as predicted theoretically ( Fig 5B ) .
In this work , we studied both theoretically and experimentally how the mechanics of cell wall expansion and the molecular processes assembling the cell wall are coordinated during cell morphogenesis , using budding yeast mating projection growth as a model system . We first derived a theoretical description of mating projection growth that couples , through a mechanical feedback encoded in the CWI pathway , the cell wall expansion and geometry to the molecular processes building the cell wall . The theoretical predictions were tested experimentally through genetic deletions affecting the feedback strength and also through mechanical perturbations ( hyposmotic shocks and cell wall degradation ) . Our theoretical predictions are in good agreement with the experimental results and indicate that the existence of mechanical feedback is essential to guarantee stability during cell wall remodeling and cell morphogenesis . This theoretical description of mating projection growth connects the mechanics of the cell wall to the molecular events in charge of sensing its mechanical state and controlling its assembly via well-established signaling pathways ( CWI pathway ) , thus providing specific predictions on how mutations can affect cell morphogenesis . Various previous models accounted for both the mechanics and assembly ( remodeling ) of the cell wall [16–19] , as we have done above , but did not account for a connection to known molecular feedback control ( CWI pathway ) coupling wall mechanics and assembly . These models consider the cell wall to be either a elastic material undergoing remodeling [17 , 18] or an elastoplastic material [16] , as opposed to our description of the cell wall as a viscous fluid , which has also been considered before [19] . Importantly , at long timescales over which cell growth and cell wall remodeling occur , assuming the cell wall to be a viscous fluid , a remodeled elastic material or an elastoplastic material is largely equivalent because all of them properly account for the observed irreversible expansion ( flow ) of the cell wall at long timescales [16] . While previous descriptions assumed that irreversible cell wall expansion only occurs when new cell wall material is inserted into the pre-existing wall [17 , 18] , we allowed the possibility of cell wall expansion even in the absence of cell wall assembly because the cell wall can be fluidized by the action of wall degrading enzymes secreted via exocytosis . Indeed , addition of zymolyase leads to cell wall piercing for cells with intact cell wall assembly ( Fig 2H ) . Such cell wall degrading enzymes are known to play an important role in cell wall remodeling [29 , 59] and the establishment of inhomogeneous cell wall material properties in several organisms [47 , 60] , including budding yeast [16] . Since these enzymes are secreted via exocytosis , we assumed the length scale of viscosity variation away from the apex to be the same as the exocytosis region . Finally , the combination of the observed inhomogeneous stiffness of the cell wall during mating projection growth [16] ( measured at short timescales; seconds ) and cell wall remodeling can be theoretically described as an effective inhomogeneous viscosity at timescales longer than cell wall remodeling , as we assumed in our description above and also done previously for other systems [19] . We theoretically find that in the absence of any mechanical feedback relaying information about the mechanical state of the cell wall to the intracellular processes building it , cell wall expansion is unstable , leading to cell lysis . Previous works have shown that the cell wall is prone to piercing in cell if the CWI is compromised [29] , and our experimental data indicates that degradation of the cell wall by zymolyase ( effectively lowering the cell wall viscosity in our description ) also leads to cell wall piercing ( Fig 2H ) . Since cell wall piercing involves changes in cell wall thickness , our theoretical description accounts for the dynamics of cell wall thickness from first principles ( mass conservation ) . This is in contrast to previous models that also consider cell wall mechanics and assembly , which assume the cell wall thickness to be constant , fixed by an unknown mechanism [16–18] . Considering a variable cell wall thickness was done before [19] , but the cell wall mechanics and assembly were considered independently ( no mechanical feedback ) and the dynamics of cell growth was not studied . We theoretically show that accounting for the simplest mechanical feedback encoded in the CWI pathway , which directly couples cell wall expansion and assembly via direct activation of Fks1/2 synthases , stabilizes cell wall expansion for a wide range of parameters . The agreement between our theoretical predictions and experimental results suggests that the specific mechanical feedback studied herein , with cell wall stress sensors Wsc1 and Mid2 locally sensing cell wall expansion and directly activating Fks1/2 cell wall synthases , can stabilize cell wall remodeling during mating projection growth by itself . Such mechanical feedback ensures that in regions where the cell wall expands the fastest ( at the projection apex ) and could potentially rupture via thinning , local activation of cell wall synthases increases assembly of cell wall material , preventing cell wall rupture and stabilizing mating projection growth . However , our work does not rule out that other mechanical feedbacks encoded in the CWI pathway could also play a role in the stabilization of projection growth . It also is likely that other stress sensors [28 , 29] , expressed during different cell wall remodeling events in budding yeast , coordinate cell wall expansion and assembly in other morphogenetic processes . While our experimental observations qualitatively agree with our theoretical predictions regarding the existence of an instability associated to the thinning of the cell wall and then need of a mechanical feedback to coordinate cell wall extension and assembly , further experiments will be needed to fully confirm this scenario . Beyond budding yeast , many other organisms , including other fungi , plants and bacteria , have walled cells that are constantly remodeled [5 , 9 , 61 , 62] . The molecular control of cell wall remodeling and morphogenesis differs across species , and it is therefore likely that different mechanisms encode mechanical feedback in other species . Indeed , previous observations have hinted at the existence of mechanical feedback [63] , but the feedback mechanisms remain elusive . The mechanical feedback described herein , or different feedback mechanisms to be discovered , may also play an important role in the coordination of cell polarity and morphogenesis in both animal and walled cells [63–66] . While essential to ensuring stability during cell wall expansion , our results show that the strength of mechanical feedback does not affect mating projection shape or size ( Figs 3 and 5 ) . The observed decoupling in the control of cell geometry and growth stability reported here may allow cells to maintain a functional shape under different growth conditions . In addition , we find that projection size is controlled by the spatial extent of exocytosis . This is in agreement with recent observations in fission yeast indicating that the size of the apical growth domain correlates best with the size of the apical exocytosis region [17] , and also with theoretical models of fission yeast that predict the radius of the cell to be determined by the size of the apical cell wall assembly region [18] . More generally , the need to coordinate growth processes and mechanics during morphogenesis is important for individual cells , but also for tissues and organs . Identifying the molecular mechanisms enabling this coordination at different scales and in different organisms will substantially contribute to our understanding of morphogenetic processes . | All morphogenesis processes , whether at the cell scale or tissue level , require the coordination of growth and mechanics to properly shape functional structures . However , the mechanisms that coordinate these two processes in the sculpting of individual cells , and especially in walled cells , remain unknown . Using yeast mating projection growth as a model system , we show that a genetically-encoded mechanical feedback relays information about the mechanical state of the cell wall to the intracellular processes assembling it , thereby coordinating cell wall expansion and growth during cell morphogenesis . We find that mechanical feedback is essential to stabilize cell growth , but the shape and size of the cell are insensitive to the feedback and independently controlled . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"cell",
"walls",
"control",
"theory",
"engineering",
"and",
"technology",
"plant",
"cell",
"biology",
"cell",
"processes",
"plant",
"cell",
"walls",
"control",
"engineering",
"viscosity",
"plant",
"science",
"developmental",
"biology",
"systems",
"science",
"mathematics",
"model",
"organisms",
"fungi",
"materials",
"science",
"experimental",
"organism",
"systems",
"materials",
"physics",
"cellular",
"structures",
"and",
"organelles",
"morphogenesis",
"chemical",
"properties",
"physical",
"chemistry",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"computer",
"and",
"information",
"sciences",
"chemistry",
"glucans",
"exocytosis",
"yeast",
"physics",
"biochemistry",
"polysaccharides",
"plant",
"cells",
"eukaryota",
"cell",
"biology",
"secretory",
"pathway",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"saccharomyces",
"cerevisiae",
"yeast",
"and",
"fungal",
"models",
"glycobiology",
"organisms"
] | 2018 | Mechanical feedback coordinates cell wall expansion and assembly in yeast mating morphogenesis |
Phenotypic screens can identify molecules that are at once penetrant and active on the integrated circuitry of a whole cell or organism . These advantages are offset by the need to identify the targets underlying the phenotypes . Additionally , logistical considerations limit screening for certain physiological and behavioral phenotypes to organisms such as zebrafish and C . elegans . This further raises the challenge of elucidating whether compound-target relationships found in model organisms are preserved in humans . To address these challenges we searched for compounds that affect feeding behavior in C . elegans and sought to identify their molecular mechanisms of action . Here , we applied predictive chemoinformatics to small molecules previously identified in a C . elegans phenotypic screen likely to be enriched for feeding regulatory compounds . Based on the predictions , 16 of these compounds were tested in vitro against 20 mammalian targets . Of these , nine were active , with affinities ranging from 9 nM to 10 µM . Four of these nine compounds were found to alter feeding . We then verified the in vitro findings in vivo through genetic knockdowns , the use of previously characterized compounds with high affinity for the four targets , and chemical genetic epistasis , which is the effect of combined chemical and genetic perturbations on a phenotype relative to that of each perturbation in isolation . Our findings reveal four previously unrecognized pathways that regulate feeding in C . elegans with strong parallels in mammals . Together , our study addresses three inherent challenges in phenotypic screening: the identification of the molecular targets from a phenotypic screen , the confirmation of the in vivo relevance of these targets , and the evolutionary conservation and relevance of these targets to their human orthologs .
Before the molecular biology era , pharmacological targets were typically classified by the effects of organic molecules on whole tissues [1] . Many pathways were first recognized based on phenotypic responsiveness to compounds without knowledge of underlying molecular mechanisms . Examples include the inference of the α- and β-adrenergic pathways in the 1940s [2] , the inference of the H2 histaminergic receptor [3] and of the μ , and κ-opioid receptors in the 1970s [4] , and the proposal of the 5-HT3 serotonergic receptor in the mid-1980s [5] . Although these targets were eventually characterized by molecular biology , the tissue and organism approach had the advantage that the compounds emerging from it were active on a physiologically intact tissue or organismal circuit , and directly linked functional perturbation of targets to biological effects . Phenotypic compound screens return to this classical approach to capture some of the same advantages for the discovery of molecules with systemic activity . Such screens have generally relied on high content microscopy assays in cell-based systems [6]–[8] . However , certain biological processes such as physiology and behavior are the result of integrated organism-wide processes that only manifest themselves in intact multicellular organisms . For example , as a physiological process , feeding behavior is the outcome of integration of extrinsic and intrinsic cues of food availability and energy demand and thus is best understood when studied in whole organisms . Elucidation of the neural circuits that determine feeding is a fundamental challenge in the neuroscience of energy homeostasis [9] . Small molecules that alter feeding behavior can serve as useful reagents for investigating these circuits and provide exquisite temporal control in ways not easily achieved through genetic manipulations . Given their small size and ease of manipulation , C . elegans have been used in pharmacology-based phenotypic screens [10]–[14] . These animals are also well suited for study of molecular and neural circuits that underlie food intake behavior . C . elegans feeds using peristaltic contractions of a muscular pharynx to aspirate microbes into the lumen of the intestine [15] . This pharyngeal pumping rate directly correlates with the transport of nutrients into the intestinal lumen [16] , [17] . C . elegans' central nervous system integrates signals from external cues such as food availability , food quality , and internal nutritional status to regulate feeding behavior [17]–[20] . Multiple pathways in the nervous system that are dependent on serotonin , glutamate , and neuropeptide release regulate the pharyngeal pumping rate in C . elegans [21]–[27] . Thus , C . elegans' feeding behavior is subject to regulation by some of the same physiological parameters and molecular components as those in mammals . Feeding behavior is a relatively challenging read-out for a screen-scale phenotypic effort . Therefore , we focused on surrogate phenotypes that could potentially enrich for identification of feeding regulatory compounds . From a high-content microscopy assay , we previously discovered 84 compounds , which increased or decreased Nile Red in C . elegans [11] . Nile Red is a vital dye that has been broadly used for detecting fat levels in numerous experimental systems [28]–[32] . However , its use as a read-out of fat content in C . elegans has been challenged [33]–[36] . Nevertheless , for a subset of compounds emerging from the initial Nile Red screen , effects on lipid content were further verified by other vital dyes , biochemical methods , molecular read-outs of fat content , as well as efficacy in mammalian cell-based models of adipogenesis [11] . Therefore , we hypothesized that compounds emerging from the Nile Red screen would be enriched for those that also alter feeding behavior . For all but one of the compounds , no biological targets were previously known . To predict targets for these compounds , we used chemoinformatic inference based on ligand patterns against mammalian receptors . For a subset of the predictions , we tested the compounds against the predicted mammalian targets in vitro and subsequently tested orthologs of these targets in C . elegans by chemical-genetic epistasis . We then combined the compounds with the mutants to map epistasis relationships for feeding behavior , identifying four signaling pathways previously unassociated with C . elegans feeding regulation . Together , our findings reveal that chemical screens in C . elegans lead to molecules with activity toward known human targets and highlight the utility of C . elegans for unambiguous assessment of compound–target–phenotype relationships in the context of an intact organism .
Computational chemoinformatics methods have been used to query annotated ligand–target interactions to identify novel targets for known drugs [37] and , recently , targets for ligands identified by a screen in zebrafish [38] . We used the Similarity Ensemble Approach ( SEA ) [39] to interrogate the ChEMBL database for targets , represented by their ligand sets , resembling compounds of the 84 identified in a C . elegans phenotypic screen [11] . ChEMBL annotates more than 8 , 000 , 000 ligand–target interactions for 2 , 456 targets; mostly , mammalian . SEA uses extended connectivity fingerprints to measure chemical similarity , quantified as Tanimoto coefficients ( Tc ) , between a query molecule and each target ligand . The ensemble of all pairwise Tc's for a target's ligand set to the query are summed and compared to an expectation of random association . Using statistical machinery similar to BLAST , an expectation value ( E-value ) quantifies the possibility that observed structural similarities could occur by chance . Because SEA relies on reported ligand–target interactions , it cannot predict associations for chemically novel molecules or for targets for which no ligands have been identified . Nevertheless , SEA provides a rapid and systematic approach for discovering the pharmacological relevance of C . elegans actives vis-à-vis mammalian targets with known ligands . At an E-value threshold of less than 10−5 , 79 of the 84 active compounds were associated with at least one target in ChEMBL . Most compounds were predicted to have two or more targets , with 572 distinct targets predicted overall ( Figure 1A ) . Target predictions often spanned two or more classes per compound as has been observed in GPCR pharmacological relationships [40] . We first considered the possibility that the profile of predicted targets simply recapitulates the underlying distribution of target classes in ChEMBL . This was not the case . Our ChEMBL-derived dataset is composed of 35% enzymes , 15% ion channels , 30% membrane receptors , and 20% transporters , transcription factors , and other proteins . Conversely , more than 60% of the SEA-predicted targets for the phenotypic actives were enzymes , whereas both membrane receptors and ion channels were comparatively underrepresented ( Figure 1B ) . This deviation may reflect the targets with known ligands that are relevant to metabolic phenotypes or peculiarities of the C . elegans model , a question that we will address below . To test the SEA predictions , we assayed selected molecules in vitro against their putative mammalian targets . Given the many potential ligand–target interactions ( 79 ligands on 572 targets , 1 , 024 overall pairs ) , we only tested compound–target interactions for which established assays were readily accessible . We tested 16 different compounds against 20 targets ( 21 ligand–target pairs ) ranging from G-protein coupled receptors ( GPCRs ) and nuclear hormone receptors ( NHRs ) to kinases and phosphatases ( Figure 2 , Table S1 ) . Nine compounds had significant activity at a 10 µM assay concentration against nine predicted targets , including the dopamine , tachykinin , oxytocin , and metabotropic glutamate receptors , the Flt-3 receptor tyrosine kinase , PI3-kinases , and the NHRs peroxisome proliferator activator receptor-gamma ( PPAR-γ ) and the androgen receptor ( Figure 2 ) . Significant activity was not observed for 12 target–ligand predictions in vitro ( Table S1 ) . A final ligand–target prediction did not show activity on the human receptor in vitro , but was confirmed by chemical-genetic epistasis in C . elegans ( below ) . This 43% in vitro hit rate resembles that observed for chemoinformatic linkage of human drugs to new [37] and to adverse drug reaction targets [41] . The potencies of the hits ranged from 8 . 6 nM to 9 . 8 µM in full concentration response analysis ( Figure 2 , Figures S1 , S2 , S3 ) . As expected , there was no correlation between E-value and potency , since target affinity is not considered when calculating ligand similarities . In several cases , SEA successfully predicted the overall target family but not the exact isoform . For instance , H6 and B16 were predicted to antagonize tachykinin receptor–3 ( Tkr-3 ) and metabotropic glutamate receptor 5 ( mGluR-5 ) , respectively , but instead antagonized Tkr-1 and mGluR-8 ( Figure 2 ) . Indeed , compared to other mGluR antagonists , B16 appears to be uniquely specific for the mGluR8 isoform ( Table S2 ) . To search for feeding regulatory compounds whose SEA-predicted targets could be confirmed by direct testing against mammalian targets in vitro , we focused on the compounds listed in Figure 2 . We found that B16 , H6 , F15 , and D20 each increased the pharyngeal pumping rate ( Figure 3A ) . Our standard assay for pharyngeal pumping rate is a real-time assay over a short time interval ( 10 s ) , in which contractions of the posterior bulb of the pharynx are manually counted . The rapid assay is particularly amenable to measuring pumping rate in large numbers of animals and experimental conditions . However , to alleviate concerns that the short duration of the assay may capture an unrepresentative aspect of the pumping rate , we compared its results with those obtained over a longer time course ( 60 s ) using time-lapse microscopy ( Figure S4 , Movies S1 , S2 , S3 , S4 , S5 ) . The long time course study revealed that for young , egg-laying gravid adults foraging on E . coli OP-50 lawns , the C . elegans pharynx contracts almost continuously with occasional brief pauses of less than a second , similar to that seen during short-term measurements . Importantly , the results of the long-term video measurements confirmed our short-term manual protocol that B16 , H6 , F15 , and D20 all increase pharyngeal pumping in a range of 7%–12% . The percent increase in pharyngeal pumping rate caused by these compound treatments is within the physiological range C . elegans exhibits in response to food following a fast , or that reported with animals treated with serotonin [17] , [27] , one of the best characterized modulators of C . elegans pharyngeal pumping . While reduced pharyngeal pumping could reflect deleterious effects on animal health such as a general disruption of neuromuscular junctions , increased pumping is less likely to be due to such nonspecific effects . To determine whether the feeding increasing effects of B16 , H6 , F15 , and D20 were dependent on specific developmental stages , we evaluated their effects when administered at different developmental exposure times ( Figure 3B ) . C . elegans treated with compounds beginning at the first larval stage ( L1 ) assayed as L4 animals exhibited similar percentage increases in feeding in response to B16 , H6 , F15 , and D20 as animals that were allowed an extra day under treatment then assayed as adults . L4 animals do have a basal pumping rate that is lower than adults , and this is reflected in different absolute pump counts for the two stages ( Figure 3B ) . Developmental exposure to the compounds was not required to elevate the pumping rate , since day 1 gravid adults raised without compound exposure still exhibited elevated pumping once exposed to these compounds . The feeding elevating effects of B16 , D20 , and F15 were notable within 1 h of exposure but that of H6 required ∼16 h ( Figure 3B ) . To ascertain whether the feeding increasing compounds mediated their effects through the targets identified in Figure 2 , we used genetic epistasis to test target engagement in vivo . We first examined compound D20 , which was predicted and shown in vitro to act on the human Flt-3 receptor tyrosine kinase , a member of the PDGF-β receptor superfamily that is involved in the early stages of hematopoiesis and is active in certain cancers [42] . While C . elegans do not have hematopoiesis , their genome encodes dozens of tyrosine kinase domains with sequence similarity to the human Flt-3 receptor with no one sequence being an obvious candidate ( Table S3 ) . To determine whether D20 interacted with any of these kinase domains to regulate feeding , we measured the pharyngeal pumping rates of RNAi-treated populations , treated with either D20 or DMSO as a vehicle control . We reasoned that if D20 exerts its effects through a receptor tyrosine kinase in C . elegans , inactivation of such a receptor should mimic the effects of D20 on feeding and , importantly , render pharyngeal pumping insensitive to further modulation . In contrast , combined pharmacological and genetic perturbations that act through independent pathways will exhibit additive or synergistic effects on feeding when combined . For each of 26 receptor tyrosine kinases with significant BLAST similarity to the Flt-3 receptor , we examined the effects of their gene knockdowns on pharyngeal pumping with or without D20 treatment ( Figure 4A ) . Among the 26 receptor tyrosine kinases , RNAi exposure of only one resulted in an interaction insensitive to D20 treatment—that of the VEGF-related receptor encoded by ver-3 ( Figure 4A ) . C . elegans subjected to ver-3 RNAi exhibit an elevated pharyngeal pumping phenotype relative to vector control-treated animals , thus mimicking the effects of D20 treatment ( Figure 4A ) . While some of the other kinase knockdowns elevated the feeding rate , they all remained sensitive to further modulation by D20 . For example , while F09A5 . 2 RNAi and frk-1 RNAi each caused increased pumping relative to RNAi vector control , D20 treatment further increased pumping rates of these RNAi-treated animals ( Figure 4A ) . Thus , the resistance of ver-3 mutants to further enhancement of feeding by D20 was not simply due to an upper physiological limit on the pumping rate . Because pharyngeal pumping is regulated by the nervous system and RNAi is sometimes ineffective at gene knockdown , particularly in the C . elegans nervous system [43] , it is possible that some of the intended gene products could not be sufficiently knocked down by our RNAi strategy . However , using RNAi to knock down gene expression products of each of ver-3 , ver-2 , egl-15 , and vab-1 , all of which have reported nervous system expressions ( Table S3 ) , led to effects on pharyngeal pumping that were similar to those obtained when we examined mutants in each of these genes ( Figure 4C ) . Thus , the combination of in vitro binding assays and patterns of phenotypic interactions in vivo strongly supported the notion that D20 mediates its feeding increasing effects in a ver-3–dependent mechanism . To further compare the pharmacological parallels between Flt-3 inhibitors and pharyngeal pumping , we tested a known Flt-3 receptor inhibitor ( 5′-fluoroindirubinoxime: 5-flurox , Figure 4B ) for its effects on pharyngeal pumping . D20 induces a dose-dependent increase in the pharyngeal pumping rate with an EC50 of 600 nM ( Figure 4B ) . Similarly , 5-flurox mimicked the dose-dependent effect of D20 with an EC50 of 40 nM ( Figure 4B ) . The increased in vivo efficacy correlates with the measured in vitro activities of the Flt-3 receptor for D20 ( Ki = 165 nM ) and 5-flurox ( IC50 = 15 nM ) [44] . Finally , joint D20 and 5-flurox treatment did not increase pharyngeal pumping rate beyond that seen by individual compound treatments ( Figure S5A ) . These combined results suggest that a target pharmacologically similar to the Flt-3 receptor exists in C . elegans to regulate pharyngeal pumping . To verify the RNAi results and further ascertain that D20 and 5-flurox elicit similar effects on C . elegans feeding behavior , we focused on ver-3 and its closest family members , ver-1 , ver-2 , ver-4 , and egl-15 . Similar to the RNAi results , mutants in either ver-2 or ver-3 exhibited elevated rates of pharyngeal pumping , while those of ver-1 and ver-4 resembled wild-type , and egl-15 mutants had reduced rates of pumping ( Figure 4C ) . As with D20 , the elevated feeding rates of ver-3 mutants were insensitive to further increase with 5-flurox treatment ( Figure 4C ) . In contrast , treatment of egl-15 , ver-1 , and ver-4 receptor mutants with either D20 or 5-flurox led to similar increases in pumping rate despite different basal pumping rates of these mutants . Treatment of the ver-2 mutants , whose basal pumping rate was increased relative to WT , returned to untreated , WT rates with either compound , recapitulating the effect observed when D20 treatment was combined with ver-2 ( RNAi ) ( Figure 4A ) . The reason for this antagonistic relationship is unclear to us but may reflect a dependence of D20's feeding elevated phenotype on intact ver-2 signaling , perhaps due to a compensatory mechanism between the different receptors . These observations support the notion that the tyrosine kinase receptor VER-3 is pharmacologically orthologous to the human Flt-3 receptor and responsible for D20's feeding phenotype . SEA predictions may prove informative for finding mechanistic targets in C . elegans even if they fail to modulate the predicted mammalian targets . This appears to be the case for K9 , an analog of D20 also predicted to inhibit Flt-3 kinase ( Table S1 , Figure 4C ) . K9-treated C . elegans resembled D20-treated animals in the dose-dependent increase in feeding rate observed ( Figure 4D ) . In addition , K9 exhibited the same genetic interactions as D20- and 5-flurox–treated animals: wild-type , egl-15 , ver-1 , and ver-4 mutants all exhibited elevated pharyngeal pumping ( Figure 4C ) ; ver-3 mutants were insensitive to treatment; and the fast pumping ver-2 mutant reverted to the wild-type rate upon K9 treatment . Therefore , the similarity of the pharmacological response between the C . elegans ver-3 and the human Flt-3 receptors , while substantial , is clearly not identical . Based on the SEA and in vitro binding data , we next examined the possibility that an oxytocin receptor-like system may underlie the feeding regulatory effects of F15 ( Figure 2 ) . The oxytocin receptor system modulates mammalian feeding [45]–[47] , but the existence of a C . elegans oxytocin receptor ortholog with a role in feeding regulation has not been defined . Both F15 and a structurally independent oxytocin receptor antagonist L-371257 exhibited dose-dependent increases in C . elegans pharyngeal pumping with in vivo EC50's of 400 and 10 nM , respectively ( Figure 5A ) . Like D20 and 5-flurox , their in vivo C . elegans efficacies parallel their in vitro affinities toward the human receptor ( Oxtr Ki's for F15 , 1 . 6 µM and for L-371257 , 4 . 6 nM ) [48] . Furthermore , simultaneous treatment with F15 and L-371257 did not further elevate pumping ( Figure S5B ) , suggesting the existence of a target with pharmacological similarity to the human oxytocin receptor that regulates feeding . BLAST comparisons suggest multiple C . elegans sequences with similarity to the human oxytocin receptor . These include an oxytocin/vasopressin-like receptor NTR-1 ( T07D10 . 2 ) that modulates male mating behavior [49] and associative learning [50] , a tachykinin receptor-like protein TKR-1 , and three related receptors ( GNRR-1 , -2 , -3 ) ( Table S3 ) . We examined the pharyngeal pumping rates in ntr-1 and the gnrr-1 , -2 , and -3 mutants ( Figure 5B ) and animals treated with tkr-1 RNAi ( Figure 5C ) . Both ntr-1 and gnrr-2 mutants exhibited wild-type feeding rates and were sensitive to F15- and L-371257-induced pharyngeal pumping increases . Both gnrr-1 and -3 mutants ( Figure 5B ) as well as tkr-1 ( RNAi ) animals ( Figure 5C ) exhibited elevated pumping in the absence of compound treatment . However , only the pharyngeal pumping rate of the gnrr-1 mutants was resistant to further increase by F15 and oxytocin-antagonizing L-371257 treatments ( Figure 5B , C ) . Turning to the tachykinin system , each of wild-type , ntr-1 , gnrr-1 , -2 , and -3 mutants responded with increased feeding on treatment with SB222200 , a high-affinity human tachykinin receptor antagonist ( Figure 5C ) . In contrast to their F15 sensitivity , tkr-1 ( RNAi ) –treated animals were indeed insensitive to the pharyngeal pumping rate increases elicited by either SB222200 or H6 treatment ( Figure 5C ) , a compound identified in our C . elegans phenotypic screen and shown in vitro to act on a human tachykinin receptor ( Figure 2 ) . Wild-type H6-dosed animals were insensitive to SB222200-induced feeding increase ( Figure S5C ) consistent with their in vitro activities as tachykinin receptor antagonists . Together these results indicate that a tachykinin-like and an oxytocin-like receptor pathway function in parallel to regulate pharyngeal pumping in C . elegans . Despite relative similarities in sequence and feeding phenotype , gnrr-1 and tkr-1 mutants were differentiated by their antagonists and pharmacologically linked to the human oxytocin and tachykinin receptors , based on responsiveness to F15/L-371257 and H6/SB222200 , respectively . Finally we tested whether B16 , the mammalian mGluR-8 inhibitor , regulates the pharyngeal pumping rate through any of the C . elegans metabotropic glutamate receptors . The C . elegans genome encodes three orthologs of human mGluR-8: mgl-1 , -2 , and -3 [51] . These receptors are expressed in the C . elegans nervous system and regulate diverse aspects of C . elegans behavior and physiology , but have not previously been implicated in feeding [52] . We found that mgl-2 but not mgl-1 mutant animals exhibited an elevated pharyngeal pumping rate in the absence of B16 treatment ( Figure 5D ) . The elevated feeding of mgl-2 mutants resembled that of WT animals treated with B16 , and that of animals treated with MMPIP [53] , a human mGluR-7 allosteric antagonist ( IC50's , 26–220 nM ) structurally distinct from B16 ( Figure 5D ) . Combined treatment of wild-type animals with B16 and MMPIP resulted in no further increase over either alone ( Figure S5D ) . The pharyngeal pumping rate of mgl-2 mutant animals , unlike both WT and mgl-1 mutant animals , was unaffected by treatment with either B16 or MMPIP , consistent with the notion that both agents mediate their feeding phenotype though inhibition of MGL-2 . Similar to WT and mgl-1 mutants , mgl-2 mutant animals further elevated pumping when treated with the Oxtr/GNRR-1 antagonist F15 ( Figure 5D ) . The additive effect of F15 on the mgl-2 mutants' pharyngeal pumping rate distinguishes its underlying biological mechanism from that of the mGluR/mgl-2 antagonists . To further examine the in vivo specificity of the compound-induced feeding phenotypes and determine whether each target regulates feeding independently of one another , we examined the feeding phenotypes of all possible binary combinations of compounds with the high pumping mutants identified by this study ( Table S4 ) . The resulting 54-interaction matrix ( Figure 6A ) classifies interactions between compounds and mutants based on the sensitivity of a mutant to a compound's effect on pharyngeal pumping . Considering that relative to wild-type animals on vehicle control each of the gene knockdowns and compound treatments individually were sufficient to cause feeding increases , two patterns of interactions were expected: interactions where the feeding increasing effects are additive , representing likely parallel mechanisms of actions , and those that are nonadditive , suggesting a single regulatory pathway . Pharyngeal pumping rates that exceeded those observed in the vehicle-treated mutant controls are classified as additive . For example , both H6 and SB222200 increase the pharyngeal pumping rates of mgl-2 , ver-3 , gnrr-1 , ver-2 , and gnrr-3 mutants , a series of additive interactions , but not that of animals treated with tkr-1 RNAi , a genetic inactivation in their common target . We also noted 10 interactions in which elevated pumping rates of specific mutants were lowered upon compound treatment . For instance , while each of D20 , K9 , and 5-flurox treatments alone , as does inactivation of ver-2 , elevate feeding rates in wild-type and in many mutant backgrounds , treatment of ver-2 mutants with each of these compounds lowers feeding ( Figure 4C , Figure 6A ) . Similar antagonistic patterns were also seen , for example , when ver-3 mutants were treated with the Oxtr antagonists F15 and L-371257 . While the precise reasons for antagonistic interactions are not known , one likely possibility is that distinct signaling pathways normally act in compensatory manners that are revealed by the simultaneous inhibition of both of these pathways . Strikingly , we noted that in all cases where compounds share a common target , for example F15 , L-371257 , and the oxytocin receptor , both compounds exhibited identical interactions , unique to each compound pair across the mutant series . This includes additive , nonadditive , and antagonist interactions . This result strongly supports the hypothesis that the compounds share a common in vivo target identified by this study that drives the feeding phenotype . Beyond specificity , this matrix also indicates a higher level pathway organization of these mutants ( Figure 6A ) . Both antagonists of the mgl-2 pathway ( B16 and MMPIP ) and VER-3 inhibitors ( D20 , K9 , and 5-flurox ) interact nonadditively and reciprocally with ver-3 and mgl-2 mutants , indicating that these gene products may act in a common pathway . However , the additive effect of B16/MMPIP versus the antagonistic phenotypic effect of D20/K9/5-flurox activity on gnrr-1 mutants is consistent with their targets being distinct entities . In addition , GNRR-1 antagonists ( F15 and L-371257 ) interact nonadditively with ver-2 mutants and antagonistically with ver-3 , recapitulating the antagonistic interaction of VER-3 inhibitors with ver-2 mutants ( Figure 4C , Figure 6A ) . This indicates that a second feeding regulatory pathway combines GNRR-1 and VER-2 signaling . To evaluate whether the chemical-genetic epistasis interactions observed were the result of pharmacological peculiarities or could be confirmed by standard genetic epistasis tests , we tested genetic interactions in the implied mgl-2/ver-3 and ver-2/gnrr-1 pathways . The feeding rate of ver-3 mutants was insensitive to effects of mgl-2 ( RNAi ) , but was reduced by ver-2 ( RNAi ) ( Figure 5B ) . Mgl-2 ( RNAi ) interacted additively with gnrr-1 mutants to increase the pumping rate , and ver-3 ( RNAi ) antagonized the feeding increasing effects of the gnrr-1 mutation ( Figure 5B ) . Ver-2 ( RNAi ) interacted nonadditively with the gnrr-1 mutants and was antagonistic to mgl-2 mutants ( Figure 5B ) . Similar interactions between gnrr-1 mutants and ver-3 ( RNAi ) , ver-2 ( RNAi ) , and mgl-2 ( RNAi ) were observed by measuring pumping over longer intervals by time lapse microscopy ( Figure S6 ) . These results indicate that the chemical-genetic interactions observed in this study accurately predict the interactions of loss-of-function perturbation combinations on the pharyngeal pumping rate . As in chemical-genetic interactions , examination of mutant combinations also indicated that mgl-2/ver-3 and ver-2/gnrr-1 function in parallel pathways but with significant crosstalk in regulating the pharyngeal pumping rate .
Whole organism phenotypic screens retain key advantages of classical pharmacological approaches , such as the discovery of compounds that are biologically active and that alter physiologically intact , integrated circuits without predisposed conceptions as to which circuits should be targeted . To prove biologically informative , this forward pharmacological approach requires the determination of in vivo molecular targets as well as the mode of action by which the phenotype is modulated [54] . Five key observations emerge from this study . First , by chemoinformatic inference , targets may be rapidly prioritized for experimental testing on isolated receptors in vitro . Whereas this method did not always succeed , the confirmation of the predicted targets was high enough , at 43% , to be practical . Second , as the identified targets are overwhelmingly mammalian , the ease of phenotypic screening strategies in C . elegans can be linked to identification of human-relevant targets . Third , in a model system such as C . elegans , the relationships of orthologous targets to in vivo phenotypes can be parsed by applying the rationale of genetic epistasis analysis . This is critical for unambiguous in vivo establishment of mechanisms of action , as in vitro activities of even highly characterized compounds are only suggestive of the in vivo efficacy targets . Fourth , despite significant differences in primary sequence identity of the targets , the in vivo efficacy in C . elegans can reflect the in vitro activity against human targets . Finally , the chemical-genetic interactions described in this study illuminate four previously uncharacterized , parallel molecular targets that regulate food intake . Together , these findings demonstrate an experimental and computational path from phenotypic screens in C . elegans to the discovery of human-relevant targets and elucidation of mechanisms of actions of newly identified compounds in C . elegans . Most drugs interact with multiple targets in vivo [41] , which can confound the assignment of phenotypic effects to particular targets . This can be especially true of compounds emerging from screening campaigns prior to any efforts aimed at optimizing the potential specificities of compounds . An advantage of a pharmacological approach in C . elegans is that genetic perturbations can test target engagement in vivo . Identifying a chemical-genetic epistatic interaction does not imply that a given compound has absolute specificity for a particular target in a biological system . However , the epistasis interaction does confirm that the compound induces a particular phenotype specifically through its interaction with the pathway defined by the genetic perturbation . Thus , in a manner similar to classical double mutant analysis , chemical and genetic interactions are combined to interrogate the pharmacological relevance of hypothesized target interactions to specific phenotypes . Our findings suggest that there is a substantial pharmacological intersection between mammals and C . elegans . These results indicate that a C . elegans phenotypic screen can lead to identification of compounds that are sufficiently similar to the mammalian pharmacopeia to allow for prediction and confirmation of their interactions with mammalian targets . In turn , evolutionary conservation of these targets , in both sequence and ligand recognition , makes it possible to accurately predict the C . elegans target whose perturbation results in the phenotype . It could have easily been the case that the compounds emerging from a C . elegans screen are so diverse as to belie prediction of targets , and that the C . elegans phenotypes could be irrelevant to the human target space , or could reflect new targets not previously seen . As such , it may be astonishing that this approach worked at all . In fact , 79 of 84 active compounds could be chemoinformatically linked to human targets , suggesting that even a diversity library retains substantial and , for our purposes , highly useful biases towards previously “liganded” targets . Moreover , whereas these in vitro mammalian targets need have no relevance for C . elegans in vivo pharmacology , for compounds with confirmed activity in vitro , orthologous targets were indeed found to mediate their C . elegans phenotypes . Whether the specific compounds identified from the C . elegans screen act on mammalian feeding and regulatory systems remains to be determined . However , there are already some hints that functionally related circuits modulate feeding behavior in both mammals and C . elegans . For instance , several independent studies in chickens , mice , and rats indicate that administration of oxytocin reduces food intake [55] . Oxytocin appears to be a target of satiety signals since a lipid-related signal , oleoylethanolamide , positively requires intact hypothalamic oxytocin signaling to mediate its anorexigenic effects [45] . Conversely , hyperphagia associated with a high-fat diet requires synaptotagmin-4–mediated suppression of oxytocin vesicle exocytosis [46] . Furthermore , a key function of neurons of the hypothalamus that stimulate feeding involves the inhibition of a separate population of oxytocin neurons [47] . Thus , analogous to C . elegans gnrr-1 , signaling through the oxytocin receptor is a negative regulator of food intake in mice . Similarly , central administration of substance P , a TKR ligand , inhibits feeding in chicks [56] , consistent with the negative feeding regulatory role of the TKR-1 in C . elegans . While the Flt-3 receptor has no known role in the mammalian nervous system , closely related growth factor receptors , such as the platelet-derived growth factor–β ( PDGF-β ) receptor , are expressed in the hypothalamus , and administration of PDGF-β depresses food intake and anti-PDGF-β antibodies elevate food intake in rats [57] . This resembles our observation that loss-of-function in a C . elegans receptor tyrosine kinase ( ver-3 ) through either mutation or pharmacological inhibition elevates pharyngeal pumping . These targets may thus have ancient evolutionary origins in the regulation of feeding behavior . Key weaknesses of our approach merit discussion . First , whereas it is comforting that we can predict targets for most biologically active synthetic compounds even in a “diversity” library , this also reflects the restricted chemical-target space in which the field is working . A library composed of genuinely novel chemotypes might be more likely to illuminate unprecedented targets , a widely desired goal of the field . On the other hand , any such truly diverse library risks missing that small part of chemical space that is relevant to terrestrial biology , the bias towards which is , after all , a pragmatic advantage of the current libraries [58] . Second , even within this restricted ligand–target space , the chemoinformatic linkage was far from perfect , and about half of the tested compounds remain unlinked to predicted targets . While the inferential computational approach cannot replace experiment , it is a rapid , comprehensive , and quantitative assessment of biologically active small molecules with unknown mechanism . These predictions generate testable hypotheses with regards to in vivo function . Third , while for feeding , the in vitro and in vivo data provided a compelling case for mechanisms of actions of F15 , H6 , B16 , and D20 , it remains to be determined whether these compounds also act on as-of-yet undetermined molecular pathways to alter other biological processes in C . elegans . Finally , while the link between compounds , targets , and C . elegans phenotypes now seems strong for several of the active compounds , linkage between C . elegans and mammalian in vivo pharmacology remains to be drawn . In summary , whereas chemoinformatic linkage retains important liabilities , its success rate here and in earlier studies [37]–[39] , [41] is high enough to be pragmatic for target hypothesis testing . Similarly , despite critical differences between C . elegans and mammals , some of them target-based , some biology-based , the targets and ligand networks for a substantial number of small molecules are conserved enough to allow target and phenotypic association across phyla . We envision that such chemical-genetic epistasis maps could be extended to saturation mapping of the pharmacological target networks underlying feeding regulation and other processes in C . elegans . The amenability of C . elegans to genetic manipulation and pharmacological screening may find broad utility as a means to identify new small molecules with interesting phenotypes and human-relevant targets .
I10 , G7 , A5 , and L15 were purchased from Chembridge . A15 , D20 , K9 , G6 , L10 , H6 , and F14 were purchased from SPECS . B16 , N10 , and F15 were purchased from Princeton Biomolecular Research . J16 was purchased from TimTec . MMPIP hydrochloride , SB222200 , 100 nM L-371257 , and 5-fluoroindirubinoxime were purchased from Tocris Biosciences . All other chemicals were purchased from Sigma . We computationally screened 84 phenotypically active compounds against molecular target panels from the ChEMBL database ( http://www . ebi . ac . uk/chembl ) using the Similarity Ensemble Approach ( SEA ) [37] , [39] operating on 1 , 024-bit folded Scitegic ECFP_4 fingerprints [59] and Tanimoto coefficients as previously described [37] . For the target panel , we first used ChEMBL_7 ( released November 11 , 2010 ) and later moved to an updated 2 , 482 molecular target panel derived from ChEMBL_11 ( released August 9 , 2011 ) . We filtered reference ligands by molecular weight ( ≤1 , 000 Da ) and by reported affinity ( ≤10 µM ) , and then subjected all ligand structures to cleaning , standardization , and de-duplication as before [39] . Tachykinin , ghrelin , and calcium-sensing receptor activity were measured using a cell-based Ca flux assay by Multispan , Inc . ( Hayward , California ) . Cholescystokinin A and B receptors were assayed for effects on cAMP production , and nicotinic acid receptor activity was measured in a cell-based assay of forskolin-stimulated cAMP production by Multispan , Inc . Calcium-sensing receptor , PP2A phosphatase activity assays were performed by CEREP ( Celle l'Evescault , France ) using PP2A from human erythrocytes . Phospholipase C from Bacillus cereus was assayed by CEREP , Inc . using glycero-phosphatidyl ethanolamine as a substrate , monitoring diacyl glycerol production . Radioligand binding assays for PPAR-α , γ , δ and the androgen receptor were performed by CEREP , Inc . CARNA Biosciences ( Kobe , Japan ) performed in vitro kinase activity assays using recombinant catalytic domains and measured phosphorylation of an Src-derived peptide for Flt-3 and phosphatidyl inositol for PI3KCA . Radioligand binding assays for the oxytocin receptor , D4 dopamine receptor , cannabinoid CB2 receptor , and angiotensin type I and II receptors were performed as described previously , as were cell-based activity assays for mGluR 1a , 2 , 4 , 5 , 6 , and 8 [60]–[62] . Strains containing mgl-1 ( tm1811 ) X , mgl-2 ( tm355 ) I , gnrr-2 ( tm4867 ) V , and gnrr-3 ( tm4152 ) X were obtained from the National Bioresource Project for the Nematode courtesy of Dr . S . Mitani at Tokyo Women's Medical University School of Medicine . Strains containing gnrr-1 ( ok238 ) I , ntr-1 ( ok2780 ) I , egl-15 ( n484 ) X , ver-1 ( ok1738 ) III , ver-2 ( ok897 ) I , ver-3 ( ok891 ) X , and ver-4 ( ok1079 ) X were obtained from the C . elegans Genetics Center , which is funded by the NIH National Center for Research Resources ( NCRR ) . N2 ( Bristol ) strain was utilized as a reference wild-type strain . Strains were outcrossed 4 times to the wild-type background . Unless described otherwise , strains were cultivated on NGM-agar plates seeded with E . coli OP-50 as described [63] . Synchronized first larval stage C . elegans were cultured for 3 d at 20°C on a lawn of HT115 E . coli induced with IPTG to express double-stranded RNAi as described [64] . Compounds at 1 , 000× stock concentrations ( 0 . 1–10 mM ) in DMSO or an equal volume of DMSO were diluted in a suspension of E . coli OP-50 ( 100 µl of a 3× concentrated overnight culture in LB broth ) and absorbed as a single drop onto 3 . 5 cm plates containing 2 . 5 ml of NGM agar forming a well-defined lawn of bacteria . For developmental exposures , 20–30 synchronized first larval stage C . elegans derived from alkaline hypochlorite treatment of gravid adults were applied to the preseeded plates and cultured until assay at mid-L4 stage ( 2 d , 20°C ) or as day 1 gravid adults ( 3 d , 20°C ) . For naïve adult exposures , synchronized first larval stage animals were cultured on NGM-agar plates seeded with E . coli OP-50 for 3 d at 20°C , then transferred to assay plates , and pumping was assayed 1 to 16 h later . Comparisons between vehicle-treated and compound-treated animals were always performed between animals at the same developmental stage and same exposure time . Pharyngeal pumping was counted by live observation at 115–200× magnification using a stereo microscope and recorded at 10 s intervals using a manually controlled digital cell counter . Alternatively , C . elegans pumping was recorded for longer intervals ( 30–60 s ) using bright-field time-lapse microscopy at 120× magnification with a 20 images per second acquisition rate . The resulting movies were analyzed manually at a playback rate of 10 images per second . Significance was determined by one-way ANOVA applying a Dunnett's posttest when comparing multiple treatments to a single control and a Bonferroni posttest when comparing the more than two treatments against one another . For pairwise nonrepeated measured comparisons , a student's t test was used . To express data as a percentage of control , the pumping rates of compound-treated animals were divided by the mean pumping rate of the DMSO-treated wild-type animals measured in the same experiment , unless indicated otherwise . Error bars on the control samples indicate the variation around that mean , which is utilized in all statistical calculations . | Many beneficial pharmacological interventions were first discovered by observing the effects of perturbation of intact biological systems by small organic molecules without a priori knowledge of their targets . This forward pharmacological approach has the advantage of directly identifying new pharmacological agents that are active on complex biological processes . However , because of experimental feasibility , systematic application of this approach is generally limited to small animals such as the roundworm C . elegans and zebrafish , raising the question of whether use of these animals could identify compounds that act on ortholgous mammalian targets . A significant challenge in addressing this question is the determination of the molecular identities of the compounds' targets responsible for the desired phenotypic outcomes . Here we describe a computational approach for target identification based on structural similarities of newly identified compounds to known ligand interactions with mostly mammalian targets . For several of the compounds emerging from a C . elegans phenotypic screen , we predict and confirm mammalian targets using in vitro binding assays . Using genetic and pharmacological assays , we then demonstrate that a subset of these compounds alter C . elegans feeding rates through the C . elegans counterparts of the predicted mammalian targets . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | In Silico Molecular Comparisons of C. elegans and Mammalian Pharmacology Identify Distinct Targets That Regulate Feeding |
Identifying viral mutations that confer escape from antibodies is crucial for understanding the interplay between immunity and viral evolution . We describe a high-throughput approach to quantify the selection that monoclonal antibodies exert on all single amino-acid mutations to a viral protein . This approach , mutational antigenic profiling , involves creating all replication-competent protein variants of a virus , selecting with antibody , and using deep sequencing to identify enriched mutations . We use mutational antigenic profiling to comprehensively identify mutations that enable influenza virus to escape four monoclonal antibodies targeting hemagglutinin , and validate key findings with neutralization assays . We find remarkable mutation-level idiosyncrasy in antibody escape: for instance , at a single residue targeted by two antibodies , some mutations escape both antibodies while other mutations escape only one or the other . Because mutational antigenic profiling rapidly maps all mutations selected by an antibody , it is useful for elucidating immune specificities and interpreting the antigenic consequences of viral genetic variation .
Host immunity drives the evolution of many viruses . For instance , potent immunity against influenza virus is provided by antibodies against hemagglutinin ( HA ) , the virus’s most abundant surface protein [1] . Unfortunately , these antibodies also select amino-acid substitutions in the HA of human seasonal influenza A virus at a rate of over two per year [2 , 3] . This rapid evolution degrades the effectiveness of anti-influenza immunity , and is a major reason why humans are repeatedly re-infected over their lifetimes . Extensive antigenic variation is also a hallmark of several other medically relevant viruses , most prominently HIV . Efforts to induce immunity to such viruses must therefore account for antigenic variation , either by targeting vaccines against current circulating viral strains [4 , 5] or developing methods to administer [6 , 7] or elicit [8 , 9] antibodies that recognize a broad range of strains . An important component of these efforts is identifying which viral mutations escape neutralization by specific antibodies . The classic approach for identifying such mutations is to select individual viral mutants that are resistant to neutralization by antibodies . For instance , escape-mutant selections with a panel of monoclonal antibodies were used to broadly define major antigenic regions of influenza HA [10–12] . However , each such selection typically only identifies one of potentially many mutations that escape an antibody , with a strong bias towards whichever mutations happen to be prevalent in the initial viral stock . Therefore , escape-mutant selections provide an incomplete picture of the ways that a virus can escape an antibody . Another approach is to individually test antibody binding or neutralization for each member of a panel of viral variants . However , there are ∼104 single amino-acid mutants to a 500-residue viral protein , so individually creating and testing all of them is a daunting task . Therefore , even the most ambitious such studies limit themselves to a small fraction of the possible point mutations , such as by only testing mutations to alanine [13–15] . But as the current work will underscore , the antigenic effect of mutating a residue to one amino acid can be poorly predictive of the effects of mutating the same residue to another amino acid . Furthermore , the difficulty in individually generating and testing large numbers of viral variants means that such studies often use simpler assays ( e . g . , hemagglutination-inhibition , pseudovirus neutralization , or protein binding ) that can be imperfect surrogates for how well a mutation enables a replication-competent virus to escape antibody neutralization [16–18] . A complete structural definition of the interface between an antibody and antigen can be obtained using methods such as X-ray crystallography . However , obtaining such structures remains non-trivial , particularly since viral surface proteins are often heavily glycosylated [19] and sometimes conformationally heterogeneous [20] . In addition , structural definitions do not reveal which mutations actually escape antibody neutralization . Mutations at only a subset of the residues in the antibody-antigen interface actually disrupt binding [21–24] , a “hot spot” phenomenon observed in protein-protein interfaces more generally [25–27] . Here we use massively parallel experiments to rapidly map all single amino-acid mutations to HA that enable influenza virus to escape from four neutralizing antibodies . Our approach involves imposing antibody selection on virus libraries generated from all amino-acid point mutants of HA , and using deep sequencing to quantify the selection on every mutation in the context of actual replication-competent virus . The resulting comprehensive map of antibody escape reveals remarkable mutation-level idiosyncrasy for each antibody: for instance , at many residues only some of the possible amino-acid mutations confer escape , and two antibodies targeting the same residue elicit unique profiles of escape mutations . Mutational antigenic profiling therefore enables complete and high-resolution mapping of viral antibody escape mutations .
To quantify the selection that neutralizing antibodies exert on all single amino-acid mutations to a viral protein , we developed the mutational antigenic profiling strategy shown in Fig 1 . A library of viruses is generated that contains all amino-acid point mutants of the protein that are compatible with viral replication . This library is incubated with or without a neutralizing antibody , and then used to infect cells . Deep sequencing of cellular RNA measures the frequency of each mutation among the viral variants that infect cells in the presence or absence of antibody , with molecular barcoding used to increase the sequencing accuracy . We quantify the differential selection for each mutation as the logarithm of its enrichment in the antibody-treated virus library relative to the no-antibody control , and display these data as in Fig 1B . In the analysis that follows , we only consider mutations with positive differential selection . We applied mutational antigenic profiling to influenza HA . HA is a 565-residue glycoprotein that forms homo-trimers on the virion surface that are responsible for both receptor-binding and membrane fusion [1] . Current influenza vaccines are designed to induce antibodies against HA , and the strains that compose these vaccines are chosen annually with the goal of matching the antigenicity of their HAs with those in circulating influenza variants [2–5] . We chose to focus on the HA of an H1N1 strain ( A/WSN/1933 ) that was isolated from humans early in the study of influenza and then serially passaged in the lab . Our reason for choosing this strain is that classic escape-mutant selections have extensively characterized the antigenicity of closely related HAs [11 , 12] , enabling us to compare our results to those obtained using more traditional methods . The first step in mutational antigenic profiling is creating virus libraries ( Fig 1A ) . A number of techniques have recently been described to create all amino-acid point mutants of a gene in the context of a plasmid [28–30] . The last few years have also seen the description of libraries of replication-competent virus mutants generated by adapting plasmid-based viral reverse-genetics systems to accommodate libraries of mutagenized plasmids [31–35] . We utilized virus libraries created by melding these two techniques to create influenza viruses carrying all HA amino-acid point mutations compatible with viral replication [31 , 32] . We initially selected these libraries with a monoclonal antibody ( H17-L19 ) targeting the Ca2 antigenic region of HA [11] . We performed three biological replicates using independently generated virus libraries , as well as a technical replicate with one of the libraries ( Fig 2A ) . The rationale for performing biological and technical replicates was to evaluate noise arising both from variability in the virus libraries and stochasticity in the antibody selections . In each replicate , the antibody exerted strong selection for mutations at a handful of sites , and little selection on the rest of HA ( Fig 2B ) . Fig 2C shows the selection for individual amino-acid mutations in a short region in HA containing most of the epitope . Visual inspection reveals consistent selection across technical and biological replicates . Statistical analysis confirms that the site differential selection is strongly correlated among replicates ( Fig 2D ) . We next asked how the differential selection depended on the concentration of antibody used . Fig 2 shows the results of mutational antigenic profiling at an antibody concentration where the virus libraries retained only 0 . 3% of their total infectivity . We performed additional experiments using dilutions of antibody that spanned a 20-fold range . Fig 3A shows the selection at each antibody concentration . As expected , there is minimal selection when comparing replicate no-antibody controls . At progressively higher antibody concentrations , differential selection increases at most sites in the epitope , while noise at other sites remains similar to the no-antibody control . However , the increase in differential selection with antibody concentration is not entirely uniform across sites ( Fig 3A ) . Fig 3B shows that despite these complexities , the sites of greatest differential selection are similar across concentrations , indicating that the identification of escape mutations does not strongly depend on antibody concentration within the 20-fold range tested here . Prior studies have shown that sub-neutralizing doses of mixtures of antibodies can select for mutations that increase the avidity of the virus for host cell receptors , as opposed to antigenic mutations within antibody epitopes [36–38] . The range of H17-L19 concentrations tested here is likely above the range of sub-neutralizing concentrations that have been used in the past to select for avidity-enhancing mutations , and it is also possible that mixtures of antibodies targeting different epitopes promote selection for avidity mutants . Understanding the determinants of how a mutation’s differential selection depends on antibody concentration is an interesting area for future work . Overall , these results confirm that mutational antigenic profiling reproducibly identifies the HA mutations that confer escape from the monoclonal antibody H17-L19 . The identified sites of selection are robust across replicate virus libraries and concentrations of antibody . We next extended the mutational antigenic profiling to three more antibodies . We performed selections with each antibody at concentrations at which the virus libraries retained 0 . 1 to 0 . 4% of their infectivity ( S1 Table ) . Each antibody exerted strong selection at a small number of residues in HA . Fig 4A shows site differential selection across HA , while Fig 4B uses logo plots to show detailed mutation-level selection at some key positions in the antibody epitopes . We again performed three full biological replicates with each antibody , and the results were again highly reproducible among replicates ( S1 Fig ) . For each antibody , the sites of strongest differential selection were clustered in surface-exposed patches on HA’s structure that are presumably within the antibody-binding footprint ( Fig 4C and S2 Fig ) . The four antibodies target three antigenic regions: H17-L19 targets Ca2 , H17-L10 targets Ca1 , and H17-L7 and H18-S415 both target Cb [11 , 12] . As expected , H17-L19 and H17-L10 exert strong selection on entirely distinct sets of residues , but H17-L7 and H18-S415 exert selection on similar sets of residues in the Cb antigenic region . For three of the antibodies , the strongly selected residues are within short contiguous stretches of primary amino-acid sequence , but for H17-L10 the strongly selected residues are distributed across 70 residues of HA’s primary sequence . Overall , these results show that mutational antigenic profiling can comprehensively identify the selection imposed by diverse antibodies . The results above were obtained using experiments that examined tens of thousands of viral variants in parallel . How do these high-throughput measurements compare to the antigenic effects of mutations measured by traditional low-throughput methods ? To address this question , we tested some of our key findings with neutralization assays on individual viral mutants . To do this , we used site-directed mutagenesis to introduce single amino-acid mutations into the HA gene , generated viruses by reverse genetics , and performed GFP-based neutralization assays [39] . A clear observation from the mutational antigenic profiling is that at some residues , only a few of the possible amino-acid mutations are strongly selected by any given antibody , concordant with prior work showing that a limited number of mutations are sufficient for antigenic drift [40] . For instance , at HA residue 154 , the H17-L19 antibody exerts strong selection only for mutations H154E and H154D , both of which introduce a negatively charged amino acid ( Figs 4B and 5A; residues are numbered sequentially beginning at the N-terminal methionine , other numbering schemes are in S1 File ) . We generated viruses carrying the H154E mutation or a mutation to alanine ( H154A ) , which mutational antigenic profiling did not find to be under differential selection . Neutralization assays confirmed that the H154E mutant completely escaped at all antibody concentrations tested , while the H154A mutant was as sensitive to antibody as wild-type ( Fig 5A ) . Therefore , a more limited method such as alanine scanning would not have identified residue 154 as a site of escape mutations . This finding demonstrates the importance of assaying all amino-acid mutations if the goal is to comprehensively map sites of escape . Another example of mutation-level sensitivity is HA residue 148 , where antibody H17-L19 only selects for mutations to serine and threonine ( Fig 4B ) . Both the V148T and V148S mutations introduce a motif ( N-X-S/T ) that potentially leads to glycosylation of the asparagine at site 146 . To confirm that only some mutations at site 148 enable escape , we generated the V148T mutant as well as another mutant ( V148R ) that does not introduce a glycosylation motif . As expected , V148T dramatically reduced the virus’s sensitivity to the antibody , whereas V148R only had a small effect ( Fig 5A ) . The mutational antigenic profiling suggests similar mutation-level sensitivity in escape from antibody H17-L10 . At residue 234 , there is strong differential selection only for mutations to the positively charged amino-acid residues lysine and arginine ( Fig 4B ) . We generated a virus carrying one of these mutations ( P234K ) as well as a virus carrying another mutation at the same residue ( P234V ) that was not under differential selection . Neutralization assays confirmed that the P234K mutation escaped H17-L10 , while the P234V mutation caused no change in antibody sensitivity ( Fig 5B ) . Interestingly , in HA’s structure , site 234 is on a neighboring protomer relative to all the other mutations strongly selected by H17-L10 ( S2 Fig ) . Our finding that escape mutations from H17-L10 cross the HA trimer interface is consistent with the fact that this antibody only recognizes trimeric HA [41] . Escape mutations at such epitopes are discernible because mutational antigenic profiling uses actual viruses that display intact HA; such conformational epitopes might not be properly displayed in the modified forms of viral glycoproteins often used in other high-throughput methods such as phage and yeast display . Overall , these results indicate the power of mutational antigenic profiling to map residues where only a few specific amino-acid mutations lead to escape from antibody . Because this approach examines HA in its native context on influenza virions , it can comprehensively map escape mutations even in complex conformational epitopes . Two of the antibodies used in our study ( H17-L7 and H18-S415 ) target the same antigenic region of HA , with residue 89 under strong selection from both antibodies ( Fig 5C and 5D ) . Do these antibodies select the same or different amino-acid mutations at this residue ? The mutational antigenic profiling suggests that both antibodies select mutations to negatively charged amino acids ( P89D and P89E; Fig 5C and 5D ) . However , each antibody also selects a unique set of additional mutations , such as P89Y for H17-L7 and P89T for H18-S415 . We generated viruses containing the P89D , P89Y , or P89T mutations and tested their sensitivity to both antibodies using neutralization assays . In agreement with the mutational antigenic profiling , the P89D mutant escaped both antibodies , but P89Y only escaped from H17-L7 and P89T only escaped from H18-S415 ( Fig 5C and 5D ) . Thus , when two antibodies target the same site , there can be both common and antibody-specific routes of escape . Characterizing antibody escape at the level of protein sites therefore only provides a partial picture of antigenicity . A complete understanding of escape requires consideration of every mutation at every site . The antigenicity of H1 HA was originally characterized in classic experiments that selected individual viral escape mutants with a panel of mouse monoclonal antibodies [11 , 12] . These experiments identified a handful of mutations that ablated binding by each antibody ( S2 Table and underlined residues in Fig 4B ) . All four antibodies used in our study are from the original panel used in the classic experiments . We expected that the sites of differential selection identified by mutational antigenic profiling would include the previously identified mutations . Indeed , there is strong overlap between sites identified by mutational antigenic profiling and sites of mutations selected in the classic experiments ( Fig 4B ) . However , we also identified numerous additional escape mutations at those and other sites . In some cases , the sites of strongest differential selection were not identified at all in the classic experiments . For instance , as shown in Fig 4 , the classic escape-mutant selections failed to identify site 157 for H17-L19 , site 89 for H17-L7 , and site 89 for H18-S415 . Differences in the virus strains used ( as discussed below ) may account for some of these discrepancies . Additionally , it is likely that mutations at these sites were not uncovered in escape-mutant selections because each such selection only finds one mutation , with a strong bias towards those that arise from single-nucleotide changes that are prevalent in the viral stock . In contrast , our approach simultaneously examines all amino-acid point mutations . The exception to the concordance between mutational antigenic profiling and classic escape-mutant selections is antibody H18-S415 ( Fig 4B ) . The classic selections failed to identify any mutations at site 89 despite the fact that mutational antigenic profiling finds by far the strongest differential selection at this residue . This discrepancy is not due to spurious signal in the mutational antigenic profiling , since Fig 5D validates that mutations at site 89 potently escape H18-S415 . Perhaps the stochasticity of escape-mutant selections caused the classic experiments to fail to probe mutations at site 89 . It is worth noting that the differential selection exerted by H18-S415 in our experiments is substantially noisier than the differential selection for the other antibodies ( Fig 4A , S1 Fig ) . In another recent study , H18-S415 selected an escape virus containing both a mutation in HA and a mutation in the neuraminidase ( NA ) gene that decreased NA protein expression , leading to increased virus avidity for host cell receptors [42] . The selection of avidity-enhancing mutations has been observed in selection of escape viruses using mixtures of antibodies [37 , 38] , and it is even possible that the H18-S415 hybridoma cell line is not completely monoclonal . Alternatively , it is possible that it is simply more difficult for the virus to escape H18-S415 , and so there is more stochastic noise in which mutations appear in our selections . The mutational antigenic profiling also failed to find strong selection from H18-S415 for some mutations reported in the classic experiments ( L87P , S92P , and E132K; see Fig 4B , S2 Table , and S6 Fig ) . Why were these mutations selected in the classic experiments but not the mutational antigenic profiling ? An important point is that the classic experiments used a different virus strain ( A/Puerto Rico/8/1934 ) than the A/WSN/1933 strain used for our mutational antigenic profiling . In order for a mutation to be under differential selection , it must both support viral replication and affect antigenicity . The mutations L87P , S92P , and E132K are all strongly disfavored under simple selection for viral replication in the A/WSN/1933 strain [32] , which likely explains why they are not under strong differential selection in our mutational antigenic profiling . This fact is an important reminder that while mutational antigenic profiling completely maps antibody selection on all single amino-acid mutations that support viral replication in a given viral strain , it does not reveal how the effects of mutations shift with changes in strain background . It remains an open question how well measurements of the effects of mutations on viral replication [43 , 44] and antigenicity [42] can be extrapolated beyond the specific genetic backgrounds tested in the lab .
We have used a new high-throughput approach to completely map the amino-acid mutations that enable influenza virus to escape from four neutralizing antibodies . Our approach is conceptually similar to recent methods that couple deep sequencing with phage or yeast display assays for antibody binding [45–47] . But whereas those methods select for binding to antigens expressed in bacteria or yeast , our approach selects for actual neutralization in the context of replication-competent virus . Our experiments therefore measure a phenotype directly relevant to virus evolution: whether a mutation enables a virus to escape neutralization by an antibody . Our approach also bears similarities to the classic method of selecting individual viral escape mutants . However , escape-mutant selections rely on the occurrence of de novo mutations in a viral stock . Therefore , like evolution itself , such selections are stochastic , and only identify one of potentially many escape mutations . In contrast , our massively parallel experiments simultaneously examine all single amino-acid mutations , thereby minimizing stochasticity and allowing us to completely map antibody selection on all point mutations . The most striking finding from our work is the exquisite mutation level-sensitivity of antibody escape . For each of the four antibodies , we identified residues in HA where only some of the possible amino-acid mutations conferred escape . In some cases , this mutation-level sensitivity is easy to rationalize: we found examples where escape required mutations that introduce glycosylation motifs or change the charge of the amino-acid sidechain . But in other cases , the effects are not only difficult to rationalize but depend on the antibody . For instance , we identified a residue targeted by two antibodies where a mutation that escaped the first antibody had no effect on the second and vice versa . Previous studies have distinguished between an antibody’s “functional epitope” and physical footprint based on the observation that binding is disrupted by mutations at only some residues that contact the antibody [21–24] . Our findings extend this concept by showing that even within the functional epitope , only certain mutations mediate escape , consistent with the observation that a small number of amino-acid mutations in HA can cause extensive antigenic drift of H3N2 influenza virus [40] . These results underscore the shortcomings of thinking about viral antigenic evolution purely in terms of antigenic sites . For instance , many approaches to forecast and model influenza virus evolution are based on partitioning HA into antigenic and non-antigenic sites [4 , 5] . However , our work shows that for any individual antibody , it is important to consider the exact amino-acid mutation as well as the site at which it occurs . Application of mutational antigenic profiling to contemporary viral strains and antibodies will enable the prospective mapping of immune-escape mutations on a vastly more comprehensive scale than previously possible .
The influenza virus mutant libraries used have been described previously [32] . Briefly , reverse-genetics plasmids [48] encoding HA gene were mutagenized at the codon level using a previously described protocol [49] . These plasmid codon-mutant libraries were used to generate libraries of replication-competent influenza viruses using a helper-virus approach that reduced the bottlenecks associated with standard reverse genetics . The virus libraries were then passaged at low MOI to create a genotype-phenotype link between the HA protein on a virion’s surface and the gene that it carries . The viral titers in these libraries were determined by TCID50 ( 50% tissue culture infectious dose ) in MDCK-SIAT1 cells ( obtained from Sigma Aldrich ) . Three fully independent virus libraries were generated beginning with independent plasmid mutant libraries as outlined in Fig 2A . It was these low-MOI passaged virus libraries [32] that formed the starting point for the antibody selections described in the current work . The antibodies used in this study were originally isolated from mice [11 , 12] . Note that in these older papers , two different naming schemes are used for the same antibodies: H17-L19 was also called Ca3; H17-L10 was also called Ca6; H17-L7 was also called Cb15; H18-S415 was also called Cb5 . Antibodies secreted by H17-L19 , H17-L10 , H17-L7 , and H18-S415 hybridoma cell lines were purified using PureProteome A/G coated magnetic beads ( Millipore ) . The hybridomas were originally derived from mice at the Wistar Institute [11] , and were provided for this study by Scott Hensley . For the selections outlined in Fig 1 , we began by diluting each virus library in influenza growth media ( Opti-MEM supplemented with 0 . 01% heat-inactivated FBS , 0 . 3% BSA , 100 U of penicillin/ml , 100 μg of streptomycin/ml , and 100 μg of calcium chloride/ml ) to a concentration of 1 × 106 TCID50 per ml . Monoclonal antibody was also diluted in influenza growth media to a concentration twice that intended for use the selection . The virus library was then neutralized by mixing 1 ml of diluted virus with 1 ml of diluted antibody to give the final antibody concentrations listed in S1 Table . This virus-antibody mixture was then incubated at 37°C for 1 . 5 hours . No-antibody controls were “mock-neutralized” in parallel by substituting influenza growth media for the diluted antibody . At the same time , serial ten-fold dilutions of mutant virus library were made from the 1 × 106 TCID50 per ml virus stock to be used as a standard curve to measure infectivity . These dilutions represented 10% , 1% , 0 . 1% , 0 . 01% , and 0 . 001% of the 1 × 106 TCID50 dose of library used in neutralizations . The viral samples were then added to cells to allow infection by non-neutralized virions . We used MDCK-SIAT1 cells that had been plated four hours prior to infection in D10 media ( DMEM supplemented with 10% heat-inactivated FBS , 2 mM L-glutamine , 100 U of penicillin/ml , and 100 μg of streptomycin/ml ) at 2 . 5 × 105 cells per well in 6-well dishes . For the infections , we aspirated off the existing D10 media and added the 2 ml of viral sample . Duplicate infections were used for each point on the standard curve of serially diluted virus . After two hours , media in each well was then changed to 2 ml WSN growth media ( Opti-MEM supplemented with 0 . 5% heat-inactivated FBS , 0 . 3% BSA , 100 U of penicillin/ml , 100 μg of streptomycin/ml , and 100 μg of calcium chloride/ml ) after rinsing cells once with PBS to remove residual virus in the supernatant . Twelve hours later , RNA was isolated from the cells in each well using a Qiagen RNEasy Plus Mini kit by aspirating media , adding 350 μl buffer RLT freshly supplemented with β-mercaptoethanol , slowly pipetting several times to lyse cells , transferring the lysate to a RNase-free microfuge tube , vortexing for 20 seconds to homogenize , and proceeding with the manufacturer’s suggested protocol , eluting in 35 μl of RNase-free water . We estimated the percent remaining infectivity in the neutralized samples using qRT-PCR and a standard curve created using the infections with 10-fold serial dilutions of the virus libraries to give the estimates in S1 Table . For the qPCR , primers WSN-NP-qPCR-F ( 5’-GCAACGGCTGGTCTGACTCACA-3’ ) and WSN-NP-qPCR-R ( 5’-TCCATTCCTGTGCGAACAAG-3’ ) were used to amplify influenza nucleoprotein ( NP ) to quantify viral infectivity , and primers 5’-canineGAPDH ( 5’-AAGAAGGTGGTGAAGCAGGC-3’ ) and 3’-canineGAPDH ( 5’-TCCACCACCCTGTTGCTGTA-3’ ) were used to quantify canine GAPDH to correct for small differences in total RNA amounts . qRT-PCR was performed using Applied Biosystems PowerSYBR green RNA-to-Ct 1-step kit , with 40 ng of RNA in each 20 μl reaction , cycling conditions of 48°C for 30 minutes , 95°C for 10 minutes , and 40 cycles of: 95°C for 15 sec , 58°C for 1 min with data acquisition . All samples were measured in duplicate , and each assay included no-reverse-transcriptase controls . Linear regression of the relationship between the log ( infectious dose ) and the mean difference in Ct between NP and GAPDH was used to interpolate the remaining infectious dose of each antibody-neutralized sample , expressed as a percentage of the 1 × 106 TCID50 used in each neutralization . To prepare deep sequencing libraries , HA genes were amplified from the RNA isolated from infected cells by reverse transcription with AccuScript Reverse Transcriptase ( Agilent 200820 ) using HA-specific primers WSN-HA-for ( 5’-AGCAAAAGCAGGGGAAAATAAAAACAAC-3’ ) and WSN-HA-rev ( 5’-AGTAGAAACAAGGGTGT TTTTCCTTATATTTCTG-3’ ) . PCR amplification of HA cDNA and Illumina sequencing library preparation was then carried out using a previously described barcoded subamplicon sequencing protocol [32] , which was in turn inspired by the approach of Wu and coworkers [33] . The only change made to the previous protocol [32] was that in order to more effectively spread sequencing depth across samples based on the expected diversity of mutations in each sample , the number of uniquely-barcoded single stranded variants used as template for round 2 PCR was 5 × 105 to 7 × 105 for the no-antibody control samples , and 1 . 5 × 105 for the antibody-neutralized samples . Sequencing libraries with unique indices for each experimental sample were pooled and sequenced on an Illumina HiSeq2500 using 2 x 250 bp paired-end reads in rapid-run mode . S3 Table provides summary statistics of the deep sequencing libraries . The frequency of each mutation in each sample was determined by using dms_tools [50] ( http://jbloomlab . github . io/dms_tools/ ) , version 1 . 1 . 20 , to align subamplicon reads to a reference HA sequence , group barcodes to build consensus sequences , and quantify mutation counts at every site in the gene for each experimental sample . We computed the extent that each mutation is enriched by each antibody selection by comparing mutation counts in each antibody-treated sample to mutation counts from the matching no-antibody control sample , also utilizing controls to account for PCR and sequencing errors . Specifically , we compute the differential selection on each mutation as follows . The error rate ϵr , x at each site r for codon x is estimated from the apparent frequency of that mutation in our previously described sequencing of HA from wild-type plasmid using barcoded-subamplicon Illumina sequencing [32] . Specifically , the error rate was calculated as: ϵ r , x = n r , x err / ∑ y n r , y err ( 1 ) where n r , x e r r is the number of counts of codon x at site r in the wild-type plasmid sequencing library . Note that for the wildtype codon x = wt ( r ) , ϵr , wt ( r ) does not represent the rate of “errors” to this codon , but rather the fraction of reads that give the wildtype codon as expected . We then adjusted the observed counts n r , x mock and n r , x selected for codon x at site r in the mock selected and antibody selected samples , respectively , to the error-corrected counts n ^ r , x for each sample: n ^ r , x = max ∑ y n r , y n r , x ∑ y n r , y - ϵ r , x , 0 if x ≠ wt r n r , x / ϵ r , x if x = wt r . ( 2 ) This correction ignores second-order terms in which a mutant codon is incorrectly read as another mutant codon or wildtype due to sequencing errors; however , provided that both error rates and mutation rates are low ( which is the case in our experiments ) , these second-order terms can be safely ignored . To convert from codon counts to amino-acid counts , we summed the error-adjusted counts for all codons encoding each amino acid a at site r to give the error-adjusted amino-acid counts n ^ r , a mock and n ^ r , a selected for the mock selected and antibody selected samples , respectively . We then computed the relative enrichment Er , a of amino acid a at site r as E r , a = ( n ^ r , a selected + f r , selected × P ) / ( n ^ r , wt r selected + f r , selected × P ) ( n ^ r , a mock + f r , mock × P ) / ( n ^ r , wt r mock + f r , mock × P ) ( 3 ) where wt ( r ) denotes the wildtype amino acid at site r , P is a pseudocount ( set to 10 in our analyses ) , and f r , selected and f r , mock give the relative depths of the selected and mock samples at site r: f r , selected = max 1 , ∑ a n r , a selected / ∑ a n r , a mock ( 4 ) f r , mock = max 1 , ∑ a n r , a mock / ∑ a n r , a selected ( 5 ) The reason for scaling the pseudocount by the library depth is that in the absence of such scaling , if the selected and mock samples are sequenced at different depths , the estimates of Er , a will tend to be systematically different from one even if the relative counts are the same in both conditions . The mutation differential selection values are the logarithm of the enrichment values: s r , a = log 2 E r , a . ( 6 ) Mutations that confer escape from an antibody will have a larger relative frequency in the antibody-selected sample than the no-antibody control sample , and will thus have a large , positive differential selection . Therefore , we limited analysis to positive differential selection to identify antibody escape mutations . To summarize the differential selection at each site , we sum the mutation differential selection values sr , a over all amino-acids a with positive mutation differential selection and term this the positive site differential selection sr for site r: s r = ∑ a max 0 , s r , a . ( 7 ) Logoplots visualizing differential selection display each amino acid with a height proportional to the mutation differential selection sr , a . Amino acid letter codes are colored based on the physiochemical properties of the amino-acid side chain: hydrophobic ( V , L , I , M , P ) are green , nucleophilic ( S , T , C ) are orange , small ( A , G ) are pink , aromatic ( F , Y , W ) are brown , amide ( N , Q ) are purple , positively-charged ( H , K , R ) are red , and negatively-charged ( D , E ) are blue . The computer code to perform these differential selection analyses is incorporated in the dms_tools ( http://jbloomlab . github . io/dms_tools/ ) software as the program dms_diffselection . The logoplots created by dms_tools are rendered with WebLogo [51] . We performed neutralization assays using viruses carrying GFP in the PB1 segment using a previously described protocol [39] . These GFP reporter viruses were generated using seven bidirectional reverse genetics plasmids [52] encoding the PB2 , PA , HA , NP , NA , M , and NS segments of A/WSN/1933 ( kindly provided by Robert Webster of St . Jude Children’s Research Hospital ) , and a unidirectional reverse genetics plasmid pHH-PB1flank-GFP in which the coding sequence of PB1 is replaced by GFP [53] . Since these viruses carry GFP instead of PB1 , they are grown in complementing 293T-CMV-PB1 ( derived from cells purchased from the American Typed Culture Collection as described in [53] ) and MDCK-SIAT1-CMV-PB1 cells ( derived from cells purchased from Sigma Aldrich as described in [53] ) that constitutively express the WSN PB1 protein . For each HA mutation tested in the neutralization assay , the indicated amino-acid mutation was introduced into the WSN HA bidirectional reverse genetics plasmid by site-directed mutagenesis , and the HA sequence was verified by Sanger sequencing . To generate each mutant GFP-carrying virus , we transfected a co-culture of 293T-CMV-PB1 and MDCK-SIAT1-CMV-PB1 cells with the eight reverse genetics plasmids described above . For each transfection , 4 × 105 293T-CMV-PB1 and 4 × 104 MDCK-SIAT1-CMV-PB1 per well were plated in 6-well plates in D10 media four hours prior to transfection . Each well received a transfection mixture of 100 μl DMEM , 3 μl BioT transfection reagent , and 250 ng of each of the eight reverse genetics plasmids . At 20 hours post-transfection , the media was changed to WSN neutralization media , which has low autofluorescence in the GFP channel ( Medium 199 supplemented with 0 . 3% BSA , 100 U of penicillin/ml , 100 g of streptomycin/ml , 100 g of calcium chloride/ml , 25 mM HEPES , 0 . 5% FBS ) . At 72 hours post-transfection , culture supernatants were clarified by centrifugation at 2 , 000×g , aliquoted , and frozen at -80°C . The GFP-carrying viruses were titered by flow cytometry in MDCK-SIAT1-CMV-PB1 cells . For this titering , cells were plated in 12-well plates at 1 × 105 cells per well in WSN neutralization media . Four hours after plating , cells were infected with dilutions of viral supernatant . At 16 hours after infection , wells with approximately 1% of cells GFP-positive were analyzed by flow cytometry , and the fraction of GFP-positive cells was used to calculate the titer of infectious particles in each viral supernatant . For the neutralization assays , monoclonal antibody was diluted down columns of a 96-well plate in WSN neutralization media . Three replicate dilution columns were used for each virus-antibody combination . Columns without antibody were used to measure maximal fluorescence in the absence of neutralization , and columns without cells were used to measure background fluorescence in viral supernatants , which we found to contribute more background fluorescence than cells alone . The GFP reporter viruses were diluted in WSN neutralization media to 1 × 103 infectious particles per μl and 40 μl ( 4 × 104 infectious particles ) was added to each well . Plates were incubated at 37°C for 1 . 5 hours before adding 4 × 104 MDCK-SIAT1-CMV-PB1 cells to each well . After 16 hours incubation at 37°C , GFP fluorescence intensity was measured on a Tecan plate reader using an excitation wavelength of 485 nm and an emission wavelength of 515 nm ( 12-nm slit widths ) . Percent of maximal infectivity was calculated by subtracting background fluorescence signal from all wells and dividing the signal from antibody-containing wells by the signal from corresponding wells without antibody . | Many viruses evolve rapidly , and this evolution sometimes enables them to escape antibodies that would otherwise neutralize their infectivity . An important aspect of studying this evolution is determining which viral mutations can mediate antibody escape . The classic way of identifying such mutations is to select or test them one by one . However , a vast number of possible mutations can be made to a virus . For instance , there are over 10 , 000 single amino-acid mutations that can be made to the most abundant surface protein of influenza virus , hemagglutinin . This is too many to test one by one , and so all previous studies of antibody escape have examined just a fraction of the possible amino-acid mutations to any given viral protein . Here we describe a new approach to quantify the selection that an antibody exerts on all these mutations in a single experiment . This approach enables us to reproducibly and sensitively identify mutations that affect antibody neutralization—for instance , at individual sites in hemagglutinin , we can distinguish which of several different mutations have the largest effect on antibody escape . The ability to completely map viral escape from antibodies opens the door to much more detailed characterization of viral antigenic evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"organismal",
"evolution",
"medicine",
"and",
"health",
"sciences",
"microbial",
"mutation",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"influenza",
"pathogens",
"immunology",
"microbiology",
"orthomyxoviruses",
"viruses",
"mutation",
"rna",
"viruses",
"microbial",
"evolution",
"antibodies",
"monoclonal",
"antibodies",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"viral",
"replication",
"viral",
"evolution",
"biochemistry",
"point",
"mutation",
"influenza",
"viruses",
"virology",
"physiology",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"evolutionary",
"biology",
"organisms"
] | 2017 | Complete mapping of viral escape from neutralizing antibodies |
Taenia solium cysticercosis/taeniosis is emerging as a serious public health and economic problem in many developing countries . This study was conducted to determine prevalence and risk factors of human T . solium infections in Mbeya Region , Tanzania . A cross-sectional survey was conducted in 13 villages of Mbozi district in 2009 . Sera of 830 people ( mean 37 . 9±11 . 3 years ( SD ) ; 43% females ) were tested for circulating cysticerci antigen ( Ag-ELISA ) and antibody ( Ab-ELISA ) . A subset of persons found seropositive by Ag-ELISA underwent computed tomography ( CT ) scan of the brain for evidence of neurocysticercosis . Stool samples from 820 of the same participants were tested for taeniosis by copro-antigens ( copro-Ag-ELISA ) and formol-ether concentration technique . Cases of T . solium taeniosis were confirmed serologically by EITB assay ( rES38 ) . A questionnaire was used for identification of risk factors . Active cysticercosis by positive Ag-ELISA was found in 139 ( 16 . 7% ) persons while anti-cysticercal antibodies were detected in 376 ( 45 . 3% ) persons by Ab-ELISA . Among 55 persons positive for Ag-ELISA undergoing CT scan , 30 ( 54 . 6% ) were found to have structures in the brain suggestive of neurocysticercosis . Using faecal analysis , 43 ( 5 . 2% ) stool samples tested positive for taeniosis by copro-Ag-ELISA while Taenia eggs were detected in 9 ( 1 . 1% ) stool samples by routine coprology . Antibodies specifically against adult T . solium were detected in 34 copro-Ag-ELISA positive participants by EITB ( rES38 ) indicating T . solium taeniosis prevalence of 4 . 1% . Increasing age and hand washing by dipping in contrast to using running water , were found associated with Ag-ELISA seropositivity by logistic regression . Gender ( higher risk in females ) and water source were risk factors associated with Ab-ELISA seropositivity . Reported symptoms of chronic severe headaches and history of epileptic seizures were found associated with positive Ag-ELISA ( p≤0 . 05 ) . The present study indicates T . solium infection in humans is highly endemic in the southern highlands of Tanzania .
Taeniosis and cysticercosis are different stages of Taenia solium infection involving swine as intermediate hosts and humans as definitive and/or intermediate hosts . Taeniosis is the intestinal infection with the adult cestode in humans and is caused by consumption of raw or undercooked pork containing viable cysticerci of T . solium . Ingestion of infective eggs passed by a person with taeniosis either by autoinfection , direct contact with another tapeworm carrier or indirectly via ingestion of contaminated food , water , or hands may also lead to cysticercosis in humans whereby larval tapeworm cysts develop in the muscles , eye and central nervous system . Pigs get cysticercosis by ingesting T . solium eggs primarily as a result of eating feces of a human tapeworm carrier . Human cysticercosis causes a variety of neurological symptoms , most commonly seizures due to cysts in the brain , a condition known as neurocysticercosis [1] , [2] . Cysticercosis imposes substantial global burden on human beings related to epilepsy , ocular disorders , other neurological manifestations , and economic losses related to disability and lost productivity [3] , [4] . A World Health Organization ( WHO ) -commissioned systematic review of studies reporting the frequency of neurocysticercosis ( NCC ) worldwide done by [5] estimated the proportion of NCC among people with epilepsy of all ages to be 29 . 0% ( 95% CI: 22 . 9%–35 . 5% ) , indicating epilepsy is consistently associated with NCC in over one quarter of patients residing in regions where T . solium infections are endemic . In sub-Saharan Africa , people with cysticercosis have been estimated to have a 3 . 4 . to 3 . 8-fold increased risk for developing epilepsy [6] . These estimates confirm the importance of NCC infection in the etiology of epilepsy in developing countries and suggest that NCC may be associated with a very large burden in cysticercosis endemic areas where epilepsy is prevalent , that is , those countries where pork consumption occurs , pigs are managed under free-range conditions , and sanitation is poor or absent enabling pigs to have access to human feces [5] . Although the recognition of its status as a serious and emerging threat to public health in Africa is increasing [7] ( WHO , 2010 ) , the data on incidence in humans is very limited in most endemic areas due to a lack of adequate surveillance , monitoring and reporting systems [8] . In Tanzania T . solium is considered widespread in the northern , central , and southern regions based on porcine cysticercosis surveys [9] , [10] , [11] , [12] , [13] . These surveys provide initial evidence that T . solium infection is of national importance . Despite the high prevalences of porcine cysticercosis reported in different areas of the country , the presence of human T . solium infection has only been confirmed at one hospital in Mbulu district , Manyara region in the northern highlands where 13 . 7% of patients with epilepsy were detected to have probable/definitive NCC based on brain scans using computerised tomography and western blot to detect antibodies to larval T . solium [14] . However , there is still a lack of information on the T . solium situation ( both taeniosis and cysticercosis ) in the general human population [15] , thus the basis for conducting the present cross-sectional survey with the aim of establishing the prevalence and identifying risk factors associated with the transmission of human T . solium infections in rural communities in the southern highlands of Tanzania which is the main pig raising area of the country [16] .
The study was approved ( No . HD/MUH/T . 75/07 ) by the ethical committee at Muhimbili University of Health and Allied Sciences ( MUHAS ) . Permission to conduct the study in the selected villages was obtained from regional , district and local authorities . Consent for selection of a participant in a sampled household was sought from the selected individual as well as the head of the household . Before interview and sample collection , each selected participant was approached individually to obtain written informed consent . For minors ( <18 years ) informed assent to participate in the study was obtained orally , and thereafter a written informed consent for the minor's participation was signed by a parent or guardian . Participants who were to undergo CT scanning were informed about the whole procedure and were scanned only after obtaining their written consent . Participants with a history of epileptic seizures , testing positive by Ag-ELISA and/or detected with cysts in their brain were admitted to Vwawa District Hospital and treated according to standard of care [17] . Participants with taeniosis were treated with a single dose of Niclosamide 2 g orally . For ascariasis and hookworm infections , a single dose of Albendazole 400 mg orally was used while a single dose of Praziquantel 40 mg/kg was used only for participants with intestinal schistosomiasis who were seronegative by Ag-ELISA . The anthelmintic treatment was administered by the researcher ( a nurse ) under supervision of a medical doctor . No adverse effects were reported during or after anthelmintic treatment . The survey was conducted in 13 randomly selected villages in Mbozi district , Mbeya Region following a preliminary study in the district which reported a high prevalence of porcine cysticercosis ( 32% ) based on serum Ag-ELISA ( B158/B60 ) in 300 pigs in 150 households ( Eric Komba , personal communication ) . The district is located in the southern highlands of Tanzania in the southwest of Mbeya region between latitudes 7° and 9° 12′ south of the equator , longitudes 32°7′30″ and 33°2′0″ east of Greenwich Meridian with altitudes between 900–2750 meters above sea level . According to the National Bureau of Statistics ( http://www . nbs . go . tz ) , the human population of Mbozi district was 515 , 270 inhabitants in 2002 . Inhabitants mainly engage in agriculture and keep pigs as a source of income . The total number of pigs in Mbozi district in the year 2009 was estimated to be 25 , 355 ( District Veterinary Officer , Mbozi District - personal communication ) . A community-based descriptive cross-sectional study was conducted between April and July 2009 . Sample size estimation was calculated using the formula [18] , in which n = required sample size , Z = Z score for a given confidence level , P = expected prevalence or proportion , and d = precision . With an estimated prevalence of 9% for taeniosis based on a Nigerian study ( [19] , the sample size was calculated at 385 households ( one person per household ) . To adjust for potential non-compliance and design effect , the sample size was more than doubled such that 900 households were targeted during the survey . A cluster-random sampling with Probability Proportion to Size ( PPS ) technique was used for selection of households according to [20] . In each village , the principal investigator conducted a meeting and explained the purpose of the study to the local authorities and villagers , after which households for inclusion in the survey were selected on the same day . During the following days , the research team visited the selected households and identified eligible household members . Criteria of eligibility were living in the household and being between 15–60 years old . Among all eligible household members who agreed to participate in the study only one person was selected per household by simple random sampling for questionnaire administration and collection of blood and stool samples . After each person was interviewed , 10 ml of venous blood was drawn from cephalic or median cubital vein ( median basilic vein ) by medical laboratory technicians in a labelled plain blood vacutainer tube and allowed to clot at room temperature . The sera were obtained by centrifugation at 3200 rotations per min for 5 min , then aliquoted into 2 ml cryovials and stored at −20°C until tested . Similarly each participant was instructed on how to collect a stool sample and given a stool container which they returned with the sample on the same or following day . All stool samples were fixed with 10% formalin by the researcher and stored at room temperature until tested . A questionnaire based on that of the Cysticercosis Working Group of Eastern and Southern Africa ( www . cwgesa . dk/CWGESA/Action Plan/Research . aspx ) was used to collect information on risk factors and other related information from enrolled participants . The questionnaire addressed demography , pork consumption habits , hygienic and sanitary practices , presence of subcutaneous nodules and history of neurological signs . Responses on hygienic and sanitary practices were confirmed by observation method . The information regarding clinical signs was obtained by asking about the participants' previous histories of subcutaneous nodules , severe chronic headaches , loss of consciousness , epileptic seizures or partial seizures . According to the International League Against Epilepsy's Commission on Epidemiology and Prognosis definitions , epilepsy was defined as recurrent ( two or more ) epileptic seizures , unprovoked by any immediate identified cause while partial seizures were defined as presence of seizures without impairment of consciousness or awareness , i . e . maintained alertness and ability to interact with the environment . Serological analysis was carried out using a monoclonal antibody-based ELISA detecting circulating antigens of T . solium cysticerci ( Ag-ELISA ( B158/B60 ) [21] and an ELISA detecting anti-cysticercal antibodies ( Ab-ELISA ( rT24h ) [22] . Faecal studies were carried out using a copro-Ag-ELISA detecting antigens of adult Taenia species [23] . In addition , formol-ether concentration technique ( microscopic examination ) was used to detect eggs of Taenia species as well as other parasites in stool . Confirmation of T . solium taeniosis was done by detection of adult T . solium specific antibodies in sera of participants tested positive by copro-Ag-ELISA using enzyme-linked immunoelectrotransfer blot ( EITB ) assay ( rES38 ) as described by [24] . All participants with a history of epileptic seizures and seropositive by Ag-ELISA ( n = 28 ) and one-fourth of Ag-ELISA seropositive participants without a history of epileptic seizures ( n = 27 ) , selected at random , were referred to Vwawa District hospital in Mbozi district , Mbeya and then transported to Muhimbili National Hospital in Dar es Salaam to be examined for evidence of NCC using computed tomography ( CT ) scan conducted with intravenous contrast . An image series of slides was made for each patient both before and after the contrast injection which were all examined by the same radiologist . All data were initially entered into EPI info software version 6 . For analysis , the data were converted to Statistical Package for Social Scientist ( SPSS ) version 13 . Bivariate analysis was first performed by Chi-Square test to assess the marginal association between human cysticercosis seropositivity and factors such as age group , gender , religion , presence of latrine , source of drinking water , hand washing practice , pork consumption , boiling of drinking water , history of passing ploglottides , and being positive by copro-Ag-ELISA at a 95% confidence level . To assess potential associations , a multivariate logistic regression analysis was performed by calculating OR and 95% confidence intervals for human cysticercosis seropositivity ( Ag-ELISA/Ab-ELISA ) and identified risk factors . The Ag-ELISA or Ab-ELISA was entered as an outcome and the covariates that were included in the model were age group , gender , presence of latrine , source of water , hand washing practice and being positive by copro-Ag-ELISA .
The risk factors that were significant in the analysis as risks associated with human cysticercosis are presented in Table 3 . Being older age was found to be a risk for seropositivity with human cysticercosis in the regression model , as the Ag-ELISA was higher in the age groups 36–45 years ( OR = 2 . 5 ) and 46–60 years ( OR = 2 . 6 ) as compared to young people ( <25 years ) . In addition , those who reported washing their hands by the dipping method using the same water as others were more likely to be seropositive by Ag-ELISA ( OR = 3 . 8 ) . On the other hand , when fitting the risk factors for seropositivity with human cysticercosis by Ab-ELISA in the regression model , results showed the risk for seropositivity to be significantly higher for males as compared to females ( OR = 1 . 6 ) , and for people using an unsafe source of water ( OR = 1 . 9 ) . Furthermore , being confirmed a T . solium tapeworm carrier by copro-Ag-ELISA substantially increased the risk for being seropositive for cysticercosis by both tests .
Findings of the present study indicate that human T . solium infection is hyper-endemic in Mbozi district , Mbeya Region , Tanzania . Further understanding of the disease distribution and transmission burden in Tanzania is needed as well as strengthening capacity to address it . Correspondingly , cross-sectoral ( stakeholders ) control efforts at local , regional , national level as well as across endemic neighboring countries of the eastern and southern Africa region should be initiated using appropriate , sustainable approaches to decrease or eliminate the burden of the disease . | Cysticercosis caused by the zoonotic pork tapeworm , Taenia solium , is emerging as a serious public health and agricultural problem in sub-Saharan Africa . Surveys have shown cysticercosis in pigs to be highly prevalent in multiple foci in Tanzania , and a hospital-based study in the northern highlands indicated neurocysticercosis as an important cause of epileptic seizures in humans . We present here a cross-sectional community-based survey on the prevalence and risk factors of human cysticercosis and taeniosis conducted in the southern highlands – the major pig-producing area of the country . The most striking findings were that more than 15% of people surveyed were found to have active cysticercosis and nearly half of them were found to have been exposed to larval T . solium indicating a high level of environmental contamination with T . solium eggs . This was supported by finding over 4% of people having had T . solium tapeworms . A subset of persons found positive serologically for active cysticercosis underwent brain scanning and more than half of them were found to have neurocysticercosis . This strong evidence that T . solium cysticercosis/neurocysticercosis/taeniosis is highly endemic in the southern highlands of Tanzania demands urgent attention of regional and national authorities to combat the parasite . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"neurocysticercosis",
"parasitic",
"diseases",
"neglected",
"tropical",
"diseases",
"veterinary",
"science",
"infectious",
"diseases",
"infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"epilepsy",
"zoonoses",
"veterinary",
"diseases",
"zoonotic",
"diseases",
"epidemiology",
"public",
"health",
"taeniasis",
"neurological",
"disorders",
"neurology",
"cysticercosis"
] | 2013 | Prevalence and Risk Factors Associated with Human Taenia Solium Infections in Mbozi District, Mbeya Region, Tanzania |
Constraint-based modeling techniques have become a standard tool for the in silico analysis of metabolic networks . To further improve their accuracy , recent methodological developments focused on integration of thermodynamic information in metabolic models to assess the feasibility of flux distributions by thermodynamic driving forces . Here we present OptMDFpathway , a method that extends the recently proposed framework of Max-min Driving Force ( MDF ) for thermodynamic pathway analysis . Given a metabolic network model , OptMDFpathway identifies both the optimal MDF for a desired phenotypic behavior as well as the respective pathway itself that supports the optimal driving force . OptMDFpathway is formulated as a mixed-integer linear program and is applicable to genome-scale metabolic networks . As an important theoretical result , we also show that there exists always at least one elementary mode in the network that reaches the maximal MDF . We employed our new approach to systematically identify all substrate-product combinations in Escherichia coli where product synthesis allows for concomitant net CO2 assimilation via thermodynamically feasible pathways . Although biomass synthesis cannot be coupled to net CO2 fixation in E . coli we found that as many as 145 of the 949 cytosolic carbon metabolites contained in the genome-scale model iJO1366 enable net CO2 incorporation along thermodynamically feasible pathways with glycerol as substrate and 34 with glucose . The most promising products in terms of carbon assimilation yield and thermodynamic driving forces are orotate , aspartate and the C4-metabolites of the tricarboxylic acid cycle . We also identified thermodynamic bottlenecks frequently limiting the maximal driving force of the CO2-fixing pathways . Our results indicate that heterotrophic organisms like E . coli hold a possibly underestimated potential for CO2 assimilation which may complement existing biotechnological approaches for capturing CO2 . Furthermore , we envision that the developed OptMDFpathway approach can be used for many other applications within the framework of constrained-based modeling and for rational design of metabolic networks .
The reconstruction of organism-specific genome-scale metabolic models and their in silico analysis by techniques of constraint-based modeling has become a key to understand structure , function , and capabilities of metabolic networks [1–3] . Applications include the calculation of optimal flux distributions , e . g . , with respect to growth or production of certain compounds ( flux balance analysis ( FBA ) , [1 , 4 , 5] ) , exploration of the space of feasible flux phenotypes by means of pathway vectors ( e . g . , via elementary flux modes [6–8] or elementary flux vectors [9] ) , prediction of reaction/gene essentialities , integration of different types of omics data , and the identification of optimal intervention targets for rational strain design [10] . Recently , more and more efforts have been made to integrate thermodynamic information into constraint-based analysis methods [11–30] , especially into FBA-based approaches [12 , 17–21 , 25–27] and pathway-based techniques [13 , 15 , 16 , 28] . Two particular methods that have received much attention are thermodynamic FBA [17 , 26] and the Max-min driving force approach [29] . For thermodynamic FBA , additional variables for the Gibbs free energy change of the reactions together with constraints on metabolite concentrations are included in the optimization problem to identify optimal flux vectors ( and a corresponding metabolite concentration vector ) where all reactions proceed in the thermodynamically feasible direction . In contrast , the Max-min Driving Force ( MDF ) approach was proposed to determine optimal ( maximal ) thermodynamic driving forces for a given metabolic pathway [29] . If a pathway has a high MDF , then a metabolite concentration vector can be found where all participating reactions of the pathway have simultaneously high driving forces facilitating a high flux and/or a low enzyme requirement . Conversely , pathways with low MDF values will either have low flux or must be catalyzed by highly abundant enzymes to enable a significant flux . The MDF method was used to thoroughly analyze the thermodynamic efficiency of different pathways of the central metabolism [29] and to evaluate the potential of pathway designs for synthetic photo-electro-autotrophy [30] . Recently , the MDF method together with a pathway identification procedure was used to identify thermodynamically feasible synthetic pathways that were assembled with enzymes from different organisms [11] . However , the latter as well as the original approach for MDF computation presented in [29] necessitate that a specific pathway is given a priori . So far , no method exists that can directly identify pathways in a metabolic network with maximal MDF . One approach could be to enumerate the complete set of elementary modes ( or other pathway vectors ) followed by a subsequent computation of their respective MDF values . However , this approach is limited to medium-size network and can thus not be used for genome-scale models . For this reason we formulate herein a mixed integer linear program ( MILP ) problem that optimizes the MDF with respect to different constraints ( e . g . concentration ranges , ratio constraints , yield constraints , etc . ) without the prerequisite to define a specific reaction sequence a priori . This MILP identifies the optimal MDF value together with a corresponding pathway ( represented as a steady-state flux distribution ) . Hence , the result of OptMDFpathway is not only the optimal MDF value but also the associated pathway enabling the optimal driving force . In the second part of this work , we employ our new OptMDFpathway approach to assess the endogenous potential of Escherichia coli to fix CO2 via thermodynamically feasible pathways . The development of sustainable bioprocesses using CO2 as feedstock to produce valuable chemicals and fuels from CO2 is highly desirable as they have several advantages compared to chemical CO2 reduction [31] . For example , only mild reaction conditions are required or low-purity reactants can be used . Many autotrophic microorganisms are capable of catalyzing the reduction of CO2 at ambient conditions . They incorporate CO2 into valuable organic compounds via six naturally occurring carbon fixation pathways [32] . Many of these organisms exhibit a phototrophic lifestyle where the Calvin-Benson-Bassham cycle [33] is the most abundant CO2 assimilation pathway . However , volumetric productivities and CO2 capturing rates are relatively small for phototrophic conditions since the maximal rate of the carboxylating enzyme Rubisco is an order of magnitude lower than the average of central metabolism enzymes [34] and efforts to improve Rubisco’s kinetic parameters were not sufficiently successful so far [35 , 36] . Also the necessary provision of suitable photobioreactors increases the costs of large-scale bioprocesses based on phototrophic organisms . Hence , there is still a need for faster and more efficient bioprocesses for the conversion of CO2 into valuable products . Recent research with respect to biotechnological potential for CO2 fixation includes the design of synthetic CO2 capturing cycles like the CETCH-cycle [37] . Studies on CO2 fixation in heterotrophic organisms have also been reported recently . For example , the Calvin-Benson-Bassham cycle has been incorporated to Escherichia coli to enable the synthesis of biomass components from CO2 [38] , Generally , heterotrophic organisms have the advantage that growth and production rates are usually superior compared to the autotrophic life style . Although wild-type E . coli cannot capture CO2 for pure biomass synthesis due to limited energy and redox supply by typical carbon substrates , a net assimilation of CO2 can take place when certain products are synthesized from a given substrate . For example , yield-optimal production of succinate with E . coli using glucose as substrate could result in synthesis of 1 . 71 mol succinate for each mol of glucose consumed . This corresponds to a conversion of 6 mol carbon from glucose and 0 . 86 mol carbon from CO2 into 6 . 86 mol carbon of succinate . In fact , for any succinate yield higher than 1 . 5 mol succinate per mol glucose , a net assimilation of CO2 takes place . In the present study we will systematically analyze the endogenous potential of E . coli to assimilate CO2 heterotrophically with two common substrates , glucose and glycerol . In contrast to classical cycles and pathways of CO2 fixation in autotrophic organisms , these pathways will typically represent linear pathways from the substrate to the respective product involving carboxylating reaction steps ( Fig 1 ) . We analyze both a core model for the central metabolism ( EColiCore2 [39] ) as well as a genome-scale model ( iJO1366 , [40] ) and identify all substrate-product combinations with CO2 capturing potential in these models . Since CO2 fixation often requires overcoming high thermodynamic barriers , this kind of analysis essentially needs a method to search for pathways that are not only stoichiometrically but also thermodynamically feasible . The new OptMDFpathway method will enable us to identify genome-scale CO2 fixation pathways with reasonable driving force .
For assessing the thermodynamic feasibility of a metabolic pathway Noor et al . introduced the concept of the Max-min driving force ( MDF ) [29] . The MDF provides an upper bound for the maximal thermodynamic driving force of a given pathway . The MDF approach requires as inputs a reaction sequence ( the pathway ) together with ranges for metabolite concentrations , the standard change in Gibbs energy ΔrG′o for the participating reactions and ( optionally ) ratio constraints for some metabolite concentrations . The driving force of a single reaction is defined as the negative Gibbs free energy change of this reaction ( −ΔrG′ ) and a reaction is thermodynamically feasible if this value is positive . The driving force of a pathway can in turn be defined as the minimum of all driving forces of the involved reactions ( and a pathway is feasible if this minimum is positive ) . Hence , to maximize the driving force of a pathway , an optimization problem is formulated to identify a metabolite concentration profile that maximizes the minimum of all single reaction driving forces . In mathematical terms , this can be stated as a linear optimization problem [29]: Maximizex , BBSubjectto− ( ΔrG′o+RT⋅NTx ) ≥Bln ( Cmin ) ≤x≤ln ( Cmax ) ( 1 ) B represents the lower bound for the driving force of all reactions participating in a given pathway which is maximized thus eventually yielding the MDF ( in kJ/mol ) . ΔrG′o is a vector containing the standard change in Gibbs energy of the involved reactions , N is the stoichiometric matrix ( which includes the external metabolites ) , Cmin and Cmax are the vectors of metabolite concentration limits , x is the vector of logarithmized metabolite concentrations and RT is the product of the universal gas constant with temperature in Kelvin . The original MDF approach presented in [29] requires a predefined pathway ( or set of active reactions ) as input . In this work we deal with finding pathways ( for a given metabolic function ) with maximal MDF . Hence , the optimal pathway is not known beforehand and needs to be identified together with its ( optimal ) MDF value . In core or medium-scale networks , this can be done as follows: one enumerates all elementary ( flux ) modes ( EMs ) [6–8] , computes for each relevant EM ( e . g . , exceeding a given minimum product yield ) its MDF and finally ranks these EMs with respect to their MDF value thus yielding the pathway with maximum driving force at the top of this list . This approach is exhaustive and using EMs as pathways brings the advantage that all pathways are balanced with respect to their intermediate metabolites . Complete EM enumeration is normally not possible in genome-scale metabolic networks due to the combinatorial explosion of possible pathways . Again , we cannot simply use the entire network as input for finding a thermodynamically feasible pathway with maximum MDF since in problem ( 1 ) it is demanded that all reactions of the network are thermodynamically feasible whereas only a subset of all reactions in the network will be active in the optimal pathway . What is needed here is a method that identifies , for a desired phenotypic behavior , both the optimal MDF and a pathway that enables this MDF . Therefore , we formulate OptMDFpathway , a mixed-integer linear program ( MILP ) that is applicable to genome-scale networks and identifies , for a specified ( desired ) phenotype , balanced ( steady-state ) flux distributions that are optimal with respect to their MDF . This MILP combines the original MDF optimization problem ( 1 ) with standard constraints used in flux balance analysis ( FBA , [5] ) and ensures that the reactions which are active in the solution ( non-zero reaction rate ) are all thermodynamically feasible ( i . e . , their driving force is greater than zero for the direction in which it operates ) . The basis of the MILP is formed by the following equations: N^r=0 ( 2 ) αi≤ri≤βi ( 3 ) Dr≤d ( 4 ) The vector r contains the ( net ) reaction rates . In contrast to the stoichiometric matrix N used in Eq ( 1 ) , the matrix N^ comprises the internal metabolites only and can be obtained by removing the rows corresponding to external metabolites from N . Constraints ( 2 ) and ( 3 ) are the same as in standard FBA problems ( steady-state assumptions , flux bounds αi and βi which include non-negativity constraints for irreversible reactions ) while ( 4 ) can optionally be used to add other inequality constraints ( like yield constraints ) to specify ( desired ) phenotypes . We again quantify the driving forces of reactions as their negative change of Gibbs free energy ( cf . Eq ( 1 ) ) and collect them in vector f: fi=−ΔrGi′=− ( ΔrGi′o+RT⋅N∙ , iT⋅x ) . ( 5 ) where N∙ , iT is the transposed i-th column ( reaction ) of the ( full ) stoichiometric matrix N ( as in Eq ( 1 ) , N includes the external metabolites ) . Again , the logarithmized metabolite concentrations in x must thereby comply with the given concentration ranges: ln ( Cmin ) ≤x≤ln ( Cmax ) . ( 6 ) In a preprocessing step we determine the minimum ( fi , min ) and maximum ( fi , max ) values for each driving force fi subject to the concentration ranges ( 6 ) . Each reaction is associated with a binary variable zi that must be 1 when a flux flows through this reaction . This is achieved by the constraint ri≤zi⋅βi ( 7 ) In order for this to work it is necessary to split the reversible reactions into the forward and reverse directions and to adjust the flux bounds ( 3 ) of the separate directions accordingly . The opposite directions of a given reversible reaction always have the same absolute driving force value but with opposite signs . For a given concentration vector x , the direction with driving force greater than zero is the direction in which the net flux through that reaction flows . In order to maximize the minimal driving force of all active reactions ( the driving force of inactive reactions with zero flux is not taken into account ) the following constraints are added to the optimization problem: fi+ ( 1−zi ) ⋅Mi≥B ( 8 ) With K = max ( fi , max ) , we set Mi = K − fi , min and because B ≤ K these constraints are always fulfilled for all reactions with ri = zi = 0 . For all reactions with zi = 1 it must hold that fi ≥ B . Using the same objective function as in ( 1 ) Maximizex , r , BB ( 9 ) but this time subject to Eqs ( 2 ) – ( 8 ) ( Eqs ( 3 ) , ( 5 ) , ( 7 ) and ( 8 ) for each reaction i ) , results in a mixed-integer linear program ( MILP ) . Its solution ( x , r , B ) will deliver a ( not necessarily unique ) flux distribution r where , given the calculated concentration vector x , each active reaction has a driving force of at least B and there is no other steady-state flux distribution in the network with a higher B ( MDF ) with this property . In other words , the flux vector r found by the MILP represents a pathway with maximum MDF . If concentration ratios of certain metabolites are assumed to be fixed ( ci/cj = h ) then constraints of the form xi−xj=ln ( h ) ( 10 ) can be added to the MILP . For analyzing the endogenous CO2 fixation potential of E . coli we analyzed both a core as well as a genome-scale metabolic model . To get a comprehensive overview of all possible CO2 fixing pathways we used the iJO1366 model [40] which comprises 1805 metabolites and 2583 reactions . We studied in parallel a smaller core model of E . coli’s metabolism because it allows us ( 1 ) to fully enumerate and then assess all pathways ( EMs ) for net CO2 fixation and thereby to highlight the differences between the EM-based and OptMDFpathway-based approach for determining thermodynamically favorable pathways and ( 2 ) to focus only on well-known ( major ) pathways of the central metabolism thus excluding possibly unrealistic results due to pathways with low capacity in the genome-scale model . As core model of the central metabolism we used EColiCore2 ( ECC2 ) which is a sub-network of iJO1366 comprising 499 reactions and 486 metabolites [39 , 40] . ECC2 conserves major properties of iJO1366 but its moderate size allows for the complete enumeration of EMs . As substrates we considered glucose and glycerol . Substrate uptake fluxes were normalized to 1 mmol/gCDW/h in both models and no further flux bounds were used . To avoid that succinate is a mandatory byproduct under anaerobic conditions ( which is in conflict with experimental findings ) we allowed reaction R_nadh16tpp ( NADH dehydrogenase ) to be reversible ( as in ECC2 , see [39] ) . Whenever we consider an internal metabolite as a potential product an auxiliary excretion reaction was temporarily added before the respective calculations for this product were performed . Standard Gibbs free energy changes ΔrG′o were determined for all cytosolic reactions where either a mapping for model reaction names to KEGG-IDs [41] was available in the BIGG database [42] or where a mapping for all metabolite names to KEGG compound IDs was possible . The ΔrG′o were calculated via the component contribution method [43 , 44] for pH 7 . 0 and ionic strength of 0 . 1 by means of the eQuilibrator database and related script files [45] . ΔrG′o ( with associated uncertainties ) could be determined for 298 and 691 out of the 397 and 1272 cytosolic reactions contained in ECC2 and iJO1366 , respectively . Generally , ΔrG′o was not considered for transport reactions in both ECC2 and iJO1366 because these values depend strongly on environmental conditions in the compartments like pH , membrane potential , ionic strength and concentration ranges of external metabolites . The concentration limits for all metabolites were set to 1 μM as lower limit and 20 mM as upper limit . The concentration of CO2 was more strictly bounded to be in the range from 100 nM to 100 μM . The concentration ratios for ATP:ADP , ADP:AMP , NAD:NADH , NADPH:NADP and HCO3-:CO2 were fixed to 10:1 , 1:1 , 10:1 , 10:1 and 2:1 , respectively . The OptMDFpathway algorithm was implemented as a new function in the MATLAB toolbox CellNetAnalyzer [46] and all calculations ( including EMs as well as flux balance and flux variability analyses ) were performed with MATLAB scripts using API functions of CellNetAnalyzer [47 , 48] .
In a first step we identified all CO2 or bicarbonate ( HCO3- ) capturing reactions in both the core and the genome-scale model ( Table 1 ) . We found nine such reactions in iJO1366 ( Table 1 ) . Carbonic anhydrase is listed as one of those , but in the following we will not consider this reaction as a carbon capturing reaction as it only supports the conversion of CO2 into bicarbonate . From their stoichiometry and reversibility , the other eight reactions hold the potential to truly fix CO2 or HCO3- in iJO1366; six of them are irreversible reactions while the backward flux of two reversible reactions ( isocitrate dehydrogenase ( R_ICDHy ) and pyruvate synthase ( R_POR5 ) ) by definition allows for carbon incorporation ( Table 1 ) . By minimizing and maximizing the thermodynamic driving force separately for each of these reactions ( with the metabolite concentration ranges given in the Methods ) , we found that , in isolation , all of the above reactions except the carbamate kinase reaction ( R_CBMKr ) hold the thermodynamic potential to proceed in direction of carboxylation since the upper ( lower ) limits of the driving forces for the irreversible ( reversible ) reactions are positive ( negative ) . The carbamate kinase reaction ( R_CBMKr ) , which catalyses the first step in the urea cycle , is defined in iJO1366 such that it consumes only one mol of ATP and ammonia for synthesis of carbamoyl phosphate . However , the overall reaction is known to proceed via three separate chemical reactions where two moles of ATP for the synthesis of one molecule of carbamoyl phosphate are utilized and the relevant nitrogen substrate under physiological conditions is glutamine [49–51] . Because of the questionable stoichiometry and the thermodynamic infeasibility we neglected R_CBMKr form further analyses with iJO1366 , however , carbamoyl phosphate can still be produced by the reaction of the carbamate kinase ( R_CBPS , Table 1 ) . In ECC2 , which contained originally only reaction R_CBMKr , we replaced the latter by R_CBPS . Further , although the backward flux of the endogenous pyruvate synthase ( R_POR5 ) reaction in the genome-scale model of E . coli is thermodynamically feasible , it remains questionable under relevant physiological conditions . It has been shown in other microorganisms that this reaction can proceed in direction of carboxylation [52 , 53] , however , the corresponding cofactor for this functionality is normally ferredoxin [52] . Given that the cofactor of pyruvate synthase in E . coli is flavodoxin and not ferredoxin [54] , the feasibility of a carboxylation function under physiological conditions remains highly unlikely as the redox potential of flavodoxin seems not sufficient to support CO2 reduction . Therefore , to avoid an unrealistic assumption , we set this reaction initially to irreversible and discuss the sensitivity of the results with respect to this modification afterwards . Hence , at this point , from the nine reactions shown in Table 1 , only six reactions do contribute to the CO2 assimilation capabilities in E . coli . The core model ECC2 contains four of these six CO2 assimilating reactions ( Table 1 ) . Next , we used flux variability analysis to check for each of the above reactions whether there exist stationary flux distributions with a positive flux of the respective reaction in direction of carboxylation . We could find such a flux distribution for all of these reactions except for the isocitrate dehydrogenase reaction ( R_ICDHyr ) . Although the latter is thermodynamically feasible in both directions , there is no balanced flux distribution in either of the two models that carries a negative flux for this reaction and it can thus not be used for CO2 fixation . At this point we can thus conclude that CO2 ( or HCO3- ) incorporation in E . coli is facilitated by one of the first five reactions given in Table 1 . Metabolites whose synthesis enables CO2 assimilation should either be products of these reactions or be transformed into different compounds by subsequent reactions . However , it is still not clear whether and for which metabolites balanced flux distributions exist in the network that lead to net CO2 consumption . In a next step we therefore used classical flux balance analysis [1 , 5] ( without thermodynamic constraints except reaction reversibility ) to identify substrate-product combinations which allow net CO2 consumption . Since we are mainly interested in identifying intracellular pathways that allow for CO2 incorporation , we restrict our analyses of potential products to the set of cytosolic carbon metabolites . Periplasmic and extracellular metabolites ( which occur in almost all cases also as cytosolic species in the model ) were not considered as possible products because analyses of the thermodynamic properties for the corresponding pathways would strongly rely on assumed environmental conditions in the specific compartments like pH , membrane potential , ionic strength or feasible concentration ranges . In total , 380 cytosolic carbon metabolites of ECC2 and 949 metabolites of iJO1366 were considered as potential ( end ) products for CO2 assimilation . For each considered potential product an auxiliary excretion reaction was temporarily added and a flux optimization ( flux balance analysis; FBA ) problem formulated with maximization of the respective excretion reaction as its objective . Since the substrate uptake rate is the only applied constraint , the resulting maximized excretion rates ( normalized to the substrate uptake rate ) coincide with optimal product yields [55] . We defined the CO2 assimilation yield YCO2/CS ( normalized to molar carbon content of the substrate ) as YCO2/CS=CP*YP/S−CSCS ( 11 ) with CP and CS representing the molar carbon content of the product and substrate and YP/S as the molar product yield . Substrate-product combinations with net CO2 assimilation were identified by selecting all products for which the CO2 assimilation yield is equal or greater than 0 . 01 which guarantees that at least 1% CO2 ( with respect to molar carbon uptake ) is assimilated in the particular product of interest . Equivalent measures for CO2 fixation in terms of yield are , for example , ( i ) the fixed CO2 per mol substrate YCO2/S=YCO2/CS*CS or ( ii ) the carbon-normalized product yield YP/SC-norm=CPCS*YP/S=YCO2/CS+1 . In the core model ECC2 , we found that synthesis of 62 of the 380 cytosolic carbon metabolites ( 16 . 1% ) allows for concomitant CO2 assimilation when using glycerol as substrate ( see Table 2 ( top-ranked products ) and S1 Table ( all products ) ) . Thereof , 18 can also be synthesized from glucose ( Tables 2 and S2 ) . The higher number of possible products when using glycerol as substrate can be explained by its higher degree of reduction ( 5 . 3 ) compared to glucose ( 4 . 0 ) . Also , the carbon-normalized CO2 assimilation yields YCO2/CS are always higher when using glycerol compared to those based on glucose ( except for oxaloacetate where identical yields are observed ) . Well-known products with net CO2 fixation are the C4-metabolites of the reductive TCA cycle branch oxaloacetate , malate , fumarate and succinate as they are part of a linear reaction sequence following phosphoenolpyruvate carboxylase . These metabolites can be produced either with glucose or glycerol as substrate and allow for the assimilation of up to 0 . 33 mol CO2 per C-mol substrate metabolized ( e . g . for oxaloacetate , Table 2 ) . However , the best normalized CO2 fixation yield is possible with orotate produced from glycerol ( Table 2 ) . Its maximal product yield is 0 . 93 mol orotate per mol glycerol which corresponds to a molar carbon assimilation yield of 0 . 55 mol CO2 per C-mol glycerol . In other words , for each supplied mol of glycerol another 1 . 65 mol of CO2 are assimilated at yield-optimal conditions . In this case , CO2 accounts for 35 . 5% of all carbon atoms of the synthesized orotate . Less obvious products that can be synthesized with high CO2 assimilation yields with both substrates are e . g . aspartate and asparagine . One metabolite that allows CO2 fixation on glycerol but not on glucose is , for instance , homoserine . In iJO1366 , as many as 253 metabolites ( 26 . 7% of all 949 cytosolic carbon metabolites ) can be synthesized with net CO2 fixation with glycerol as substrate ( Tables 3 and S3 ) . Thereof , 41 can also be synthesized from glucose with concomitant CO2 fixation ( Tables 3 and S4 ) . Despite the much larger number of metabolites whose synthesis allows in principle for CO2 assimilation , the ranking of top candidates in iJO1366 is very similar to ECC2 ( Table 3 ) . The C4-metabolites oxaloacetate , orotate , and aspartate of the reductive TCA cycle branch are again the products with highest carbon assimilation yields . However , for some products , the maximum carbon assimilation yields are up to 10% higher compared to ECC2 indicating that some pathways contained in iJO1366 but not in ECC2 allow even higher CO2 assimilation . Also , some new products show up as promising candidates , for example , ( iso ) citrate . We then used flux variability analysis in both models to determine which of the carbon assimilation reactions are essential for the identified substrate-product combinations ( Table 4 ) . The PEP carboxylase reaction ( R_PPC ) is most often essential followed by the reactions of the phosphoribosylamino-imidazole carboxylase ( R_AIRC2 ) and carbamoyl phosphate synthase ( R_CBPS ) which are essential for a smaller number of products . The two reactions of the acetyl-CoA carboxylase ( R_ACCOAC ) and dethiobiotin synthase ( R_DBTS ) ( exclusively ) contained in iJO1366 are not essential for any substrate-product combination . These two reactions fix only minor amounts of CO2 when biomass components are produced and they cannot contribute to net CO2 fixation in any product . This implies that from the five reactions in iJO1366 ( three reactions in ECC2 ) where CO2 or HCO3- is defined as consumable reactant ( Table 1 ) , eventually only three reactions are accountable for E . coli’s carbon ( net ) fixation abilities . We furthermore found that the production of each metabolite requires at least one specific essential carbon assimilation reaction ( s ) meaning that all alternative production pathways for each metabolite share the same essential carboxylation reaction ( s ) ( Table 4 ) . In the core model ECC2 , for 22 of the 62 metabolites whose synthesis from glycerol allows for concomitant CO2 assimilation , the simultaneous activity of two carboxylation reactions in ECC2 is required whereas the remaining 40 require only one carboxylation reaction . With glucose as substrate , there are four products ( N-carbamoyl-L-aspartate , ( S ) -dihydroorotate , orotate , orotidine-5-P ) that require two carboxylation reactions . In iJO1366 with glucose , only one metabolite ( orotidine-5-P ) essentially requires two carboxylation reactions and 51 metabolites with glycerol , respectively . At yield optimality , the number of metabolites in iJO1366 with two mandatory CO2 assimilation reactions increases to five and 91 with glucose and glycerol as substrate . The results presented so far considered exclusively stoichiometric constraints and did not yet account for thermodynamics . Therefore , in the following we will use the concept of Max-min Driving Force ( MDF; see Methods ) to identify synthesis routes that are feasible with respect to both stoichiometric and thermodynamic constraints . For a given pathway , the MDF quantifies the maximal ( best-case ) thermodynamic driving force based on standard Gibbs free energy changes and metabolite concentration ranges . In our application , the goal was to find pathways with CO2 net fixation for substrate-product combinations where the MDF is greater than zero thus indicating principle feasibility of the respective pathway . As described in the Methods section , in the core model ECC2 we computed for each substrate-product combination the set of elementary modes ( EMs ) and identified from this set all ( stoichiometrically feasible ) pathways with CO2 net fixation ( CO2 assimilation yield YCO2/CS larger than 0 . 01 ) . By definition , for each product , the maximum CO2 net fixation previously calculated with FBA ( Table 2 ) is achieved by at least one EM . For each EM with CO2 net fixation we calculated its respective MDF to test for thermodynamic feasibility ( MDF > 0 ) . Having the complete set of EMs at hand , we can also easily identify the EM ( s ) with the maximal MDF . For glucose , 16 of the 18 identified products with stoichiometric CO2 assimilation in ECC2 are , in principle , thermodynamically feasible because there exists at least one EM for these metabolites with a positive driving force ( Tables 2 , 5 , S1 and S2 ) . In fact , we found a positive MDF for all EMs of all of these 16 products . In contrast , the previously identified metabolites asparagine and oritidine-5-phosphate must be excluded as products since the highest MDF values of their EMs for these two products are negative ( -1 . 0 , cf . Table 2 ) . With glycerol as substrate , MDF analysis of the EMs revealed that 29 from the set of 62 products ( 47% ) with stoichiometric CO2 assimilation are also thermodynamically feasible , while for 33 ( 53% ) no EM with positive MDF could be found . For the 29 feasible products , on average about the half ( 48 . 2% ) of the corresponding EMs have positive MDF values ( ranging from 4 . 1% for succinate to 98 . 7% for uracil ) while others are thermodynamically infeasible . The complete list of all possible substrate-product combinations together with their stoichiometric and thermodynamic properties is given in S1 and S2 Tables . The largest MDF for any product with CO2 fixation on both substrates is given by 8 . 6 kJ/mol ( Table 2 ) . It can be achieved with different products ( e . g . , orotate and the C4-metabolites of the TCA cycle ) using either glucose or glycerol as substrate . Optimal MDF values were always observed at suboptimal product yields for all substrate product combinations . With glycerol as substrate , only nine of the 26 in principle thermodynamically feasible products are also feasible with maximal stoichiometric CO2 assimilation yield . As already indicated above , with glucose as substrate , all yield-optimal pathways to the 16 products are also thermodynamically feasible , however , always with reduced MDF compared to the maximal MDF achievable with some minimum CO2 fixation ( Table 2 ) . It has been suggested that an MDF of 3 . 0 kJ/mol would allow for large ( net ) fluxes [29] and we found that such an MDF can be achieved for all substrate-product combinations that are in principle thermodynamically feasible . For the majority of products this threshold can be reached either at yield-optimality or at least with only slightly reduced product yields ( Table 2 ) . For six promising substrate-product combinations in ECC2 we investigated the relationship between MDF and CO2 assimilation yields in more detail ( Fig 2 ) . Although maximal MDF values usually occur at suboptimal product yields , no clear functional relationship or trend between product or CO2 assimilation yield and pathway MDF can be described . For specific yields there may exist broad ranges for the MDF values of the corresponding EMs . The opposite also holds true , a specific MDF value relates to many EMs that may span a wide range of possible CO2 assimilation yields . The most outstanding metabolites are oxaloacetate and orotate . Oxaloacetate allows for the highest carbon assimilation yield at MDF optimal conditions with both substrates . Even at maximal CO2 assimilation the corresponding maximal MDF values are as high as 7 . 5 kJ/mol and 7 . 1 kJ/mol , respectively . Orotate is another metabolite showing not only a high maximal MDF value but also a high carbon assimilation yield at MDF optimality . We finally analyzed also the number of required reaction steps for synthesizing the respective products with net CO2 fixation ( pathway lengths in Table 2 ) and found that the minimal number of cytosolic enzyme-catalyzed reactions of the feasible pathways within the top candidates is relatively low ranging from ten for producing oxaloacetate with glycerol to 29 for synthesizing dihydroorotate from glucose . On average , for 84 . 7% of these reactions thermodynamic information was available . Therefore , only a minor fraction of the involved reactions may further reduce the corresponding MDF . Since MDF analysis via exhaustive enumeration and analysis of EMs as performed in ECC2 is not possible in this large-scale model , we used our new OptMDFpathway algorithm ( Methods ) to identify , for a desired phenotypic behavior ( here: CO2 net fixation with a certain minimum yield for a given substrate-product combination ) , both the optimal MDF and a pathway that enables this MDF . In iJO1366 , 34 of the 41 stoichiometrically identified products with glucose as substrate still allow for carbon fixation if the thermodynamic constraints are taken into account . With glycerol 145 out of 253 are thermodynamically feasible ( Tables 3 , 5 , S3 and S4 ) . As in ECC2 , maximal MDF values can always be observed at suboptimal product yields for all substrate-product combinations and the largest MDF for any product on both substrates is given by 8 . 6 kJ/mol which can be achieved with different products using either glucose or glycerol as substrate ( Table 3 ) . Oxaloacetate and orotate can again be identified as the most promising candidate products for both substrates ( the pathway from glycerol to orotate is exemplarily shown in S10 Table ) . However , compared to ECC2 , the increased maximal CO2 assimilation yields observed for some products are often accompanied with smaller corresponding MDF values resulting in even negative MDF ( thermodynamically infeasible ) when glycerol is applied as substrate ( Table 3 ) . However , the majority of products can at least be produced with nearly optimal product yields via pathways supporting a MDF of 3 . 0 kJ/mol or higher ( Table 3 ) . As for ECC2 , we analyzed the relationships between product yields and the corresponding maximal MDF in more detail for the three promising candidates , orotate , oxaloacetate , and aspartate ( Fig 3 ) . For each of the six considered substrate-product combinations we iteratively increased the minimal carbon assimilation yield YCO2/CS in discrete steps from 0 up to its corresponding maximum and computed the respective maximal MDF at each step . Since the space of flux distributions is reduced with higher CO2 assimilation yields , the MDF may either remain constant or will decrease with increasing product yields . As already mentioned above , in case of glycerol the very high product yields are not supported by thermodynamically feasible pathways . However , with slightly suboptimal yields considerably large driving forces are possible . In general , the behavior of the optimal MDF with respect to the CO2-assimialtion yield is similar to ECC2 ( Figs 2 and 3 ) except for the higher maximal yield for oxaloacetate . Compared to ECC2 , we see that the minimal pathway lengths for the respective products are slightly shorter ( Table 3 ) . There were 13 substrate-product combinations within the set of top candidates that require less than 17 cytosolic enzyme-catalyzed reactions , the number of enzymes required for the cell-free CETCH cycle [37] . Even if we account for potential transport and exchange reactions , these numbers appear still realistic for biotechnological applications with a cell-free approach . Again , within the identified shortest pathways , for the vast majority ( 87 . 6% ) of all reactions thermodynamic information was available . In the work [29] , Noor et al . pointed out that the MDF can be sensitive against the pH values used for the calculation of the standard Gibbs free energy change . We therefore repeated the calculations for pH-values ranging from 6 to 8 ( with a step size of 0 . 5 ) . We found that the results are fairly robust ( see Fig 4 and S6–S9 Tables ) . For example the number of thermodynamically feasible products in iJO1366 with CO2 net fixation ( Fig 4a ) changes only slightly as there are only four metabolites ( three for glycerol and one for glucose ) that are feasible at pH 7 but become infeasible at other pH conditions . The opposite holds true for 13 metabolites ( five for glucose and eight for glycerol ) for which no thermodynamically feasible pathway exists at pH 7 but at least for one alternative considered pH condition ( see S6–S9 Tables ) . The average maximal MDF values ( for all products feasible over the whole pH range ) slightly decrease for acidic pH values and increase for basic conditions ( Fig 4b ) . However , this relationship is not generally valid for all metabolites as there are products for which the corresponding average maximal MDF decreases with increasing pH values ( e . g . aspartate or oxaloacetate ) or is maximal at pH 7 ( e . g . orotate ) . The distribution of MDF values over all EMs in ECC2 shows that there are only few distinct MDF values ( Fig 2 ) . Likewise , only a limited number of different optimal MDF-values were identified in iJO1366 ( Fig 3 ) . This suggests that there exist a finite number of distinct thermodynamic bottlenecks ( TBs ) which set upper bounds for the maximal possible thermodynamic driving force . The notion of thermodynamic bottlenecks was originally introduced in [23] to mark single reactions ( localized TBs ) or groups of reactions ( distributed TBs ) that render a flux along a given pathway thermodynamically infeasible , i . e . , where , from a given range of possible metabolite concentrations , no concentration vector can be found such that the driving force of each reaction i is positive ( -ΔrGi'>0 ) . In our application we demand instead -ΔrGi'>B but the notion of localized and distributed TBs can be directly adopted . Accordingly , a localized TB occurs if a single reaction is limiting the maximal pathway MDF because it is hitting the upper boundary value for its driving force ( for reversible reactions in the respective forward or backward direction ) . For a localized TB , the concentrations of all reactants of the respective reaction reach their maximum value and all products their minimum value ( since otherwise the MDF could be further improved ) . In contrast , in a distributed TB , several reactions constrain the pathway MDF simultaneously and the driving force of one involved reaction cannot be increased without lowering the driving force of another because the reactions share some metabolites . The concentration of these metabolites must be balanced such that the minimal driving force for all participating reactions is optimized . In [29] , thermodynamic bottlenecks hindering a higher MDF in a given pathway were identified by shadow prices of the linear MDF optimization problem . However , since OptMDFpathway is a MILP problem ( operating on a network with possibly multiple optimal pathways with maximal MDF ) this shadow price approach cannot be applied here . We therefore proceed as follows . The value of the MDF-variable B in Eq ( 8 ) is fixed to the previously calculated maximal MDF . With this background each driving force ( fi ) is separately maximized as objective in Eq ( 9 ) ( instead of B ) . The reaction ( s ) whose determined maximal fi equal the maximal MDF value previously determined for the pathway are the limiting reaction ( s ) forming the localized or distributed TB . One example of a localized TB occurring in the core as well as in the genome-scale model is given by the adenylate kinase reaction ( R_ADK1 ) which sets an upper MDF limit of 8 . 6 kJ/mol . Hence , the maximal MDF of all EMs in ECC2 or pathways in iJO1366 that comprise adenylate kinase in forward direction is given by this limit ( unless an even more stringent constraint further reduces the pathway MDF ) . The upper limit for the driving force of adenylate kinase is hit at maximal concentrations for AMP and ATP and minimal concentration for ADP ( thereby accounting for the used concentration ratio constraint of 10:1 for [ATP]:[ADP] ) . Another localized TB occurring in many MDF-optimal pathways with glycerol as substrate is given by the glycerol dehydrogenase reaction ( R_GLYCDx ) with its driving force upper limit of 7 . 5 kJ/mol . This limit is reached if the concentrations of cytosolic glycerol and NAD are at their respective maximal values and the concentrations of dihydroxyacetone and NADH at their minimal values . Finally , the localized TB with the smallest positive MDF is given by the malate dehydrogenase reaction ( R_MDH ) confirming that this reaction is a potential bottleneck for biomass or , as in our application , for product synthesis ( cf . with [29] ) . An example of a distributed TB in ECC2 for products with CO2 fixation with glucose as substrate is composed of the three glycolytic reactions R_TPI , R_GAPD , and R_FBA catalyzed by triose-phosphate isomerase , glyceraldehyde-3-phosphate dehydrogenase , and fructose-bisphosphate aldolase . If these three reactions , which were also identified in [29] as a thermodynamic bottleneck for glycolysis , are simultaneously used in a pathway , then the MDF is limited by 3 . 3 kJ/mol ( again , occurrence of other bottlenecks in the same pathway may further reduce the maximal MDF ) . This limit is caused by the shared metabolite glyeraldehyde-3-phosphate ( M_g3p_c ) which is a product of reactions R_TPI and R_FBA but a substrate of reaction R_GAPD . To achieve high driving forces for R_TPI and R_FBA , low concentrations of glyeraldehyde-3-phosphate are beneficial . Contrary , sufficient driving forces for R_GAPD can only be achieved with high concentrations of glyeraldehyde-3-phosphate . Therefore , the concentration of glyeraldehyde-3-phosphate needs to be carefully balanced to enable high driving forces for all three reactions simultaneously . Altering the ( optimal ) concentration of M_g3p_c would increase the driving force of ( at least ) one reaction but lower the driving force of another thereby reducing the overall pathway driving force . The optimal MDF values of all 62 substrate-product combinations of ECC2 are limited by only 17 different bottlenecks where the largest distributed TB limit was composed of nine reactions ( S1 and S2 Tables ) . Over all EMs of all substrate product combinations , there were only 27 different MDF values , thereof 9 with glucose and 23 with glycerol as substrate , respectively . The most abundant distributed TB with two reactions in iJO1366 is given by the simultaneous operation of triose-phosphate isomerase ( R_TPI ) together with glyceraldehyde-3-phosphate dehydrogenase ( R_GAPD ) . If both reactions occur together in a pathway , the MDF of all routes comprising these reactions cannot be higher than 4 . 5 kJ/mol because glyeraldehyde-3-phosphate is a substrate of R_TPI but a product of R_GAPD . Therefore , its concentration needs to be balanced such that both reactions simultaneously achieve the highest possible driving force . Notably , this distributed TB allows for a higher driving force compared to the three-reaction TB discussed above for ECC2 which contained additionally R_FBA and caused a lower MDF limit of 3 . 3 kJ/mol . In iJO1366 alternative reactions with favorable thermodynamic properties can substitute for R_FBA making it dispensable and allowing the higher MDF value . Setting the reaction of the pyruvate synthase ( R_POR5 ) reversible in iJO1366 ( thus allowing activity in direction of carboxylation ) increases the number of products with CO2 assimilation only slightly . However , for some products the maximal CO2 assimilation yield increases significantly . Formate is one such example: with R_POR5 being reversible , significantly higher product yields of 10 . 5 mol/mol glucose and 6 . 1 mol/mol glycerol ( compared to 6 . 14 and 3 . 38 in iJO1366 with R_POR5 being irreversible ) with accompanying molar carbon assimilation yields ( YCO2/CS ) of 0 . 75 and even 1 . 03 ( compared to 0 . 02 and 0 . 13 ) , respectively . However , as the reversibility of the endogenous R_POR5 reaction remains highly suspicious , a heterologous expression from an organism where this enzyme has been demonstrated to be fully functional in direction of carboxylation seems to be more promising . Desulfovibrio africanus is one possible organism whose respective enzyme showed high stability even under aerobic conditions . However , the corresponding cofactor ferrodoxin possibly needs to be transferred as well since E . coli does not possess cofactors whose redox potentials are sufficiently low . Not surprisingly , both metabolic models used herein predict that biomass synthesis with net CO2 fixation . is not possible in E . coli . However , a much less intuitive result found in the genome-scale is the following: if an electron source is provided that can permanently reduce NAD+ to NADH ( e . g . , via bioelectrochemical approaches [56] ) , then E . coli could grow with CO2 as the only carbon source . Clearly , under these assumptions , ATP can be produced via respiration with NADH as electron donor but this also implies that the genome-scale model must contain a cycle with net CO2 fixation . A closer look revealed that such a cycle indeed exists in the model which involves reactions of the PEP carboxylase , the TCA cycle and glyoxylate shunt as well as reactions in the serine and threonine metabolism . This cycle can produce pyruvate from CO2 and NADH ( from which then biomass can be synthesized ) . It has an MDF of 8 . 6 kJ/mol and is thus theoretically thermodynamically feasible , however , whether it would really have sufficient capacities to allow growth of E . coli solely from CO2 and a source of reduction equivalents remains to be shown . We finally note here an interesting theoretical result highlighting the relationship between the steady-state flux vectors with maximal MDF and the EMs ( which are steady-state flux vectors with minimal ( irreducible ) sets of active reactions ) . In the smaller network ECC2 , where the EMs could be enumerated , we implicitly assumed that MDF-optimal pathways with desired properties can be directly identified from the set of computed EMs . Indeed , also for genome-scale networks , it can be shown that there exists always at least one EM with maximum MDF value . The reasoning is as follows: Consider we know all EMs; this allows us to select the one with optimal MDF value . Adding further ( active ) reactions to this EM would impose further thermodynamic constraints , hence , the MDF either reduces or , in the best case , remains constant . Since removing a reaction from an EM implies that only the zero vector remains as feasible steady-state flux distribution , no flux vector in the network can exist with higher MDF . This implies that a solution found by OptMDFpathway is either an EM or it uses a superset of reactions active in an MDF-optimal EM . In the latter case , the reaction set being active in the found optimal flux vector can always be reduced to the pathway represented by the optimal EM ( where the active reversible reactions in the EM are used in the same direction as in the flux vector ) . This is a direct consequence of the fact that every flux vector can be written as a conformal sum ( sum without cancellations ) of EMs [57] . If inhomogeneous constraints are used in the model definition , the same reasoning implies that there is always an elementary flux vector ( a generalization of EMs [9] ) with optimal MDF . These findings complement other theoretical results regarding the role of EMs for the identification of optimal pathways with respect to different objectives , for example , yield-optimal pathways [55] and pathways with maximal specific rates in kinetic models [58 , 59] .
It is usually assumed that heterotrophic organisms like E . coli cannot be applied to assimilate significant amounts of CO2 . We argue that the potential of CO2 assimilation by endogenous pathways of heterotrophic organisms may have been a so far overlooked component for sustainable bioprocesses consuming CO2 . To verify this hypothesis , we systematically identified , for the first time , all combinations of two industrially important substrates and cytosolic carbon metabolites ( products ) in E . coli which lead to net CO2 fixation . By using a new optimization approach , OptMDFpathway , we ensured not only stoichiometric but also thermodynamic feasibility of the identified pathways . Our analyses complement autotrophic approaches for biotechnological CO2 assimilation by investigating CO2 fixation routes that require the constant supply of carbon substrates to which CO2 can be attached . We demonstrated that E . coli can assimilate CO2 into many different metabolites: in principle , 15 . 3% of all cytosolic carbon metabolites in the E . coli genome-scale model can be synthesized with concomitant CO2 fixation from the considered substrates via thermodynamically feasible pathways . The potential products include the expected metabolites of the left branch of the TCA cycle but also less obvious candidates . We found that 40% ( 150 of 374 cases; Table 5 ) of all substrate-product combinations with stoichiometric net CO2 assimilation in the two networks are thermodynamically infeasible emphasizing the need of a method such as OptMDFpathway to filter the high percentage of thermodynamically infeasible pathways . If glucose is used as substrate , fewer products allow for net-carbon assimilation but the relative proportion of thermodynamically feasible substrate-product combinations is higher compared to glycerol ( Table 5 ) . Although the ordering and maximal carbon assimilation yields of the top-candidates remained largely unchanged , the number of possible products more than tripled when the genome-scale model iJO1366 instead of the core model ECC2 is used and some of the maximal product yields increase more than 10% . The best substrate-product combination showing a high CO2 assimilation yield together with sufficient driving force and a small number of participating reactions was determined to be the synthesis of oxaloacetate from glycerol via glycolysis and PEP carboxylase . The optimal corresponding pathway converts one mol of glycerol into one mol of oxaloacetate , fixes one mol of CO2 and enables thus a molar carbon fixation yield of 0 . 33 . The found pathway requires only one carboxylation reaction ( PEP carboxylase ) and supports a high MDF of 7 . 5 kJ/mol . Further , the pathway comprises only 13 enzyme catalyzed reactions whereof four are needed for NADH regeneration . Opposed to the proposed cell-free CETCH-cycle [37] , this pathway ( as all pathways identified herein ) is even balanced with respect to ATP and redox cofactors such that no additional redox/energy sources must be provided . In a cell-free setup , redox cofactor regeneration could be facilitated by a bioelectrochemical transfer ( by means of suitable mediators ) of the electrons to an electrode . The proposed process would on the one hand further reduce the number of required enzymes ( to nine ) and on the other hand generate an electron flow that can be harvested to improve the overall process performance . Another interesting compound not mentioned so far which can be synthesized with a relatively high CO2 assimilation yield and with the highest molar product yield of all candidates is carbamoyl phosphate ( Table 3 ) . Carbamoyl phosphate is an industrially relevant product as it can be further processed to synthesize , for example , different cyanates from which insecticides or polyurethanes can be derived . In E . coli it is produced by the carbamoyl phosphate synthase ( Table 1 ) and functions as an intermediary metabolite for nitrogen disposal through the urea cycle and for the synthesis of pyrimidines . However , when considering it as potential sink for CO2 , a nitrogen source needs to be supplied . Clearly , the predicted thermodynamic feasibility of the identified pathways represent best-case scenarios . Whether the upper limits of the corresponding pathway driving forces can be experimentally established in vivo remains to be shown . Since we restricted the MDF maximization on the set of reactions that proceed exclusively in the cytosol , it is not per se guaranteed that the here identified pathways can be readily applied in vivo . For experimental validation of particular substrate product combinations membrane transportation costs should be carefully analyzed with respect to the specific envisioned environmental conditions . Herein we did not explicitly consider introduction of heterologous enzymes and pathways in E . coli but identified the pyruvate synthase as one target for improved CO2 assimilation capabilities . Furthermore , apart from monetary values of substrates and products , the economic viability and overall CO2 sequestration yield of a process where E . coli synthesizes one of the identified products with net CO2 assimilation will require a more detailed analysis . For example , the CO2 sequestration during ( photo ) synthesis of the respective carbon feedstock as well as the possible CO2 release for growth of E . coli must then be taken into account . The assessment of the thermodynamic feasibility of pathways with net CO2 fixation in the genome-scale model of E . coli was only possible with the development of the new OptMDFpathway method . This MILP-based algorithm determines , for a given ( e . g . , desired ) phenotype both the optimal MDF and a supporting pathway . Herein we used this approach to assess the thermodynamic feasibility of pathways in E . coli that allow for CO2 net fixation but we envision that it can as well be applied to many more general problems in metabolic network modeling and design . In particular , it can be used as a generic method to identify , in large-scale networks , pathways with desired properties ( e . g . , synthesis of a chemical with some minimum yield ) and with maximal possible driving forces . | When analyzing metabolic networks , one often searches for metabolic pathways with certain ( desired ) properties , for example , conversion routes that maximize the yield of a product from a given substrate . While those problems can be solved with established methods of constraint-based modeling , no algorithm is currently available for genome-scale models to identify the pathway that has the highest possible thermodynamic driving force among all solutions with predefined stoichiometric properties . This gap is closed with our new approach OptMDFpathway which is based on the recently introduced concept of Max-min Driving Force ( MDF ) . OptMDFpathway offers various applications , especially in the context of metabolic design of cell factories . To demonstrate the power and usefulness of OptMDFpathway , we employed it to analyze the endogenous CO2 fixation potential of Escherichia coli . While E . coli cannot assimilate CO2 into biomass , net CO2 fixation can take place along linear pathways from substrate to product and we show that thermodynamically feasible pathways with net CO2 assimilation exist for 145 ( 34 ) products when choosing glycerol ( glucose ) as substrate . Our results indicate that heterotrophic organisms like E . coli hold a possibly underestimated potential for CO2 assimilation which may complement existing biotechnological approaches for capturing CO2 . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"carbohydrate",
"metabolism",
"ecology",
"and",
"environmental",
"sciences",
"chemical",
"compounds",
"atmospheric",
"science",
"carbohydrates",
"glucose",
"metabolism",
"organic",
"compounds",
"glucose",
"monomers",
"(chemistry)",
"metabolites",
"thermodynamics",
"polymer",
"chemistry",
"proteins",
"carboxylation",
"metabolic",
"pathways",
"chemistry",
"atmospheric",
"chemistry",
"physics",
"biochemistry",
"environmental",
"chemistry",
"organic",
"chemistry",
"post-translational",
"modification",
"earth",
"sciences",
"glycerol",
"monosaccharides",
"carbon",
"dioxide",
"biology",
"and",
"life",
"sciences",
"greenhouse",
"gases",
"physical",
"sciences",
"metabolism"
] | 2018 | OptMDFpathway: Identification of metabolic pathways with maximal thermodynamic driving force and its application for analyzing the endogenous CO2 fixation potential of Escherichia coli |
LYST is a large cytosolic protein that influences the biogenesis of lysosome-related organelles , and mutation of the encoding gene , LYST , can cause Chediak-Higashi syndrome . Recently , Lyst-mutant mice were recognized to also exhibit an iris disease resembling exfoliation syndrome , a common cause of glaucoma in humans . Here , Lyst-mutant iris phenotypes were used in a search for genes that influence Lyst pathways . In a candidate gene–driven approach , albino Lyst-mutant mice homozygous for a mutation in Tyr , whose product is key to melanin synthesis within melanosomes , exhibited complete rescue of Lyst-mutant iris phenotypes . In a genetic background–driven approach using a DBA/2J strain of congenic mice , an interval containing Tyrp1 enhanced Lyst-dependent iris phenotypes . Thus , both experimental approaches implicated the melanosome , an organelle that is a potential source of oxidative stress , as contributing to the disease phenotype . Confirming an association with oxidative damage , Lyst mutation resulted in genetic context–sensitive changes in iris lipid hydroperoxide levels , being lowest in albino and highest in DBA/2J mice . Surprisingly , the DBA/2J genetic background also exposed a late-onset neurodegenerative phenotype involving cerebellar Purkinje-cell degeneration . These results identify an association between oxidative damage to lipid membranes and the severity of Lyst-mutant phenotypes , revealing a new mechanism that contributes to pathophysiology involving LYST .
LYST is a large cytoplasmic protein that influences several traits relevant to human health and disease [1] . Mutations in the encoding gene , LYST , can cause Chediak-Higashi syndrome , a rare , autosomal recessive disorder characterized by variable degrees of oculocutaneous albinism , immunodeficiency , prolonged bleeding time , and progressive neurologic dysfunction [2] , [3] . Lyst-mutant mice also exhibit ocular defects resembling exfoliation syndrome [4] , a common disease that is characterized by iris defects , fibrillar accumulations , and aberrantly dispersed pigment throughout the anterior chamber of the eye [5] . As fibrillar material and dispersed pigment accumulate in the outflow structures of the eye , intraocular pressure can become elevated and a secondary form of glaucoma often ensues . The extent to which Chediak-Higashi syndrome and exfoliation syndrome resemble each other at a mechanistic level remains to be determined , but both disease states clearly share important links to LYST . Since the time the Lyst gene was initially discovered [2] , [6] , a cellular framework for understanding LYST function has only partially emerged . LYST is present in most tissues [7] and loss-of-function mutations lead to the enlargement of lysosome-related organelles including lysosomes , melanosomes , and platelet-dense bodies [8] . In this enlarged state , the organelles often fail to undergo normal movements [9]–[12] , and exhibit altered protein components consistent with defective protein trafficking [13]–[16] as well as impaired lysosomal exocytosis leading to defects in plasma membrane repair [11] . LYST contains relatively few motifs with definitive function , thus providing limited insight into how LYST protein might contribute to these defects . Domains present in LYST include several ARM/HEAT repeats located near the amino terminus , a perilipin domain , a BEACH domain , and seven WD40 repeats located near the carboxy terminus [1] . Multiple protein-protein interactions involving LYST have been identified , including interactions with HGS , YWHAB ( commonly referred to as 14-3-3 ) , and CSNK2B [17] . Collectively , these studies suggest that LYST organizes protein-complexes important to lysosome-related organelles , perhaps through interactions with membrane domains . Here , a genetic approach for expanding knowledge of Lyst function is undertaken . The goal of these experiments is to identify genetic modifiers of Lyst-mediated phenotypes in mice . C57BL/6J mice homozygous for the beige-J mutation of the Lyst gene ( B6-Lystbg-J ) exhibit a unique iris phenotype characterized by iris stromal atrophy , pigment dispersion , dark iris color , and altered morphology of the iris pigment epithelium [4] , . Because the iris is easily assayed , we reasoned that these iris phenotypes could form a convenient basis for genetic screens of Lyst-dependent modifiers . Two approaches are taken , one candidate-based and another based on manipulation of genetic background . Both experimental approaches implicate oxidative stress as contributing to the mechanism of disease . Testing this hypothesis directly , we found that Lyst mutation leads specifically to an accumulation of lipid hydroperoxides . Likely a consequence of impaired lysosomal exocytosis and a resulting failure in plasma-membrane repair , these findings implicate oxidative membrane damage as a pathological component of Lyst-mutant phenotypes .
Previously , adult B6-Lystbg-J mice were shown to have an iris disease involving pigment dispersion and a distinct transillumination defect [4] , [18] . To determine whether these phenotypes are the consequence of altered development or an early-onset degenerative disease , iris phenotypes of B6-Lystbg-J and C57BL/6J control mice were compared throughout postnatal development ( Figure 1 ) . While the iris of C57BL/6J mice remained relatively constant with age ( Figure 1A–1L ) , the iris of B6-Lystbg-Jmice followed a degenerative course ( Figure 1M–1U; additional time points provided in Figure S1 ) . At 17 days of age , when mice became just large enough to examine with an ophthalmic slit-lamp , the iris of B6-Lystbg-J mice appeared relatively normal , by both slit-lamp examination ( Figure 1M and 1P ) and histologic analysis ( Figure 1S ) . At 60 days of age , slit-lamp analyses indicated early signs of iris disease characterized by a dark and granular-appearing iris ( Figure 1N ) . At this age B6-Lystbg-J mice also exhibited minor concentric iris transillumination defects ( Figure 1Q ) , which have previously been shown to correlate with altered morphology of the iris pigment epithelium [4] . Histologic analysis confirmed both that the iris stroma was atrophied and that the morphology of the iris pigment epithelium was altered ( Figure 1T ) . By 100–135 days of age , these changes had become more striking ( Figure 1O , 1R , and 1U ) , with the most notable change being that the iris transillumination defects were more pronounced . Collectively , these results indicate that iris disease in B6-Lystbg-J mice is the consequence of an early-onset degenerative process . Having established this , we next set out to identify genetic modifiers of these Lyst-mutant phenotypes that might shed light on the underlying molecular mechanisms . As in the case of B6-Lystbg-J mice , DBA/2J mice also develop a degenerative iris disease involving iris stromal atrophy and iris transillumination defects [19] , [20] . The iris disease of DBA/2J mice is caused by digenic interaction of two genes encoding proteins found within melanosomes , Tyrp1 and Gpnmb [21] , and can be rescued by mutations that decrease pigment production [21] , [22] . To test whether pigment production is also important to the iris disease of B6-Lystbg-J mice , genetic epistasis experiments were performed . Albino B6 . Tyrc-2J mice were intercrossed with B6-Lystbg-J mice to generate mice homozygous for both mutations on a uniform C57BL/6J genetic background ( B6 . Tyrc-2J Lystbg-J ) . The rationale for this experiment was that if pigment production contributes to Lyst-mutant phenotypes , B6 . Tyrc-2J Lystbg-J mutant irides lacking pigment production should exhibit suppressed phenotypes . Cohorts of B6 . Tyrc-2J Lystbg-J mutant mice were generated and analyzed ( Figure 2 ) . The Tyrc-2J mutation rescued all observable Lyst-mediated iris phenotypes , with B6 . Tyrc-2J Lystbg-J eyes indistinguishable from control B6 . Tyrc-2J eyes ( B6 . Tyrc-2J Lystbg-J , n = 20 eyes at 2–5 months , 12 eyes at 9–11 months , 72 eyes at 12–19 months; B6 . Tyrc-2J , n = 30 eyes at 2–5 months , 8 eyes at 9–11 months , 34 eyes at 12–19 months ) . The iris stroma of B6 . Tyrc-2J and B6 . Tyrc-2J Lystbg-J eyes were free of stromal atrophy ( Figure 2A and 2B ) , with no accumulations of macrophages or debris in the anterior chamber ( Figure 2C and 2D ) . All eyes exhibited transillumination defects typical of albino mouse eyes , with no indication of the concentric transillumination defect characteristic of Lyst-mutant mice ( Figure 2E and 2F ) . Rescue was confirmed by histologic analysis of the iris ( Figure 2G and 2H; additional time points in Figure S2 ) . Together , these results identified Tyr as a genetic suppressor of Lyst-mutant iris phenotypes , and indicated that melanin production contributes to the pathological events leading to iris disease in B6-Lystbg-J mice . To complement the candidate-driven search for potential Lyst modifiers , a genetic background-driven approach was also undertaken by creating and analyzing a congenic strain of DBA/2J mice containing the Lystbg-J mutation ( D2 . Lystbg-J ) . The rationale for this experiment was that Tyrp1 mutation , Gpnmb mutation , or other factors from the DBA/2J genetic background might affect Lyst-mutant iris phenotypes . After 10 generations of backcrossing , D2 . Lystbg-J mice homozygous for the Lystbg-J mutation were generated and assayed for relevant iris phenotypes ( Figure 3 ) . The Lystbg-J mutation caused a lightening of the DBA/2J coat color ( Figure S3 ) . At all ages examined , the DBA/2J background enhanced Lystbg-J ocular phenotypes ( n = 30 eyes of D2 . Lystbg-J mice 1–7 months of age ) . At ages when DBA/2J mice with wild-type Lyst alleles exhibited only mild indices of iris abnormalities ( Figure 3A , 3D , and 3G ) , D2 . Lystbg-J mice exhibited severe disease ( Figure 3B , 3E , and 3H ) that was enhanced over that in B6-Lystbg-J mice ( Figure 3C , 3F , and 3I ) . In D2 . Lystbg-J irides , the extent of iris stromal atrophy and iris transillumination defects was notably worsened , and resulted in large accumulations of pigment within the inferior irideocorneal angle . These results indicate that the DBA/2J genetic background enhances iris phenotypes of Lystbg-J mice . The identity of the DBA/2J modifier was subsequently shown to be located within a small region of mouse chromosome 4 and is likely the Tyrp1b mutation . DBA/2J mice have a known mutation in the Tyrp1 gene [23] , which similar to the Lystbg-J mutation , also causes iris stromal atrophy [19] , [21] . To directly test whether Tyrp1 genotype influences Lyst phenotypes in mice , a wild-type Tyrp1 allele was crossed onto the D2 . Lystbg-J genetic background by intercrosses with the previously described D2 . Tyrp1B6GpnmbB6 congenic strain of mice [24] . Irides of DBA/2J mice with differing Lyst and Tyrp1 genotypes were subsequently compared ( Figure 4 ) . As described above , the Lystbg-J mutation results in a subtle , but readily detectable , pattern of iris transillumination defects on the C57BL/6J genetic background ( Figure 4A ) , a phenotype that is greatly enhanced on the DBA/2J genetic background ( Figure 4B ) . Among 39 ( D2 . Lystbg-J X D2 . Tyrp1B6GpnmbB6 ) F2 progeny examined , a total of 11 mice exhibited transillumination defects of two severities . Four mice homozygous for the Tyrp1b mutation , but with at least 1 wild-type Lyst allele , exhibited mild transillumination defects ( Figure 4C ) . Seven mice homozygous for the Lystbg-J mutation , but with at least 1 wild-type Tyrp1 allele , exhibited moderate transillumination defects ( Figure 4D ) . The severity of transillumination defects for DBA/2J mice with wild-type Tyrp1 were greatly reduced in comparison to those in D2 . Lystbg-J mice ( compare Figure 4D to Figure 4B ) . Gpnmb genotype , which was also segregating in these crosses , had no discernable influence . Quantification based on analysis of the amount of red light present in images of these eyes ( Figure S4 ) led to the same conclusion , transillumination defects in DBA/2J mice with wild-type Tyrp1 were significantly reduced in comparison to those in D2 . Lystbg-J mice ( P<0 . 001 , Student's two-tailed t-test ) . These results map a DBA/2J-derived modifier of Lyst to a small ( approximately 14–36 cM , ref [22] ) congenic interval that encompasses the Tyrp1 gene . Because the Tyrp1b mutation is the only known mutation within this interval in DBA/2J mice , Tyrp1b is likely to be the causative modifier . The TYRP1 protein is often affiliated with an enzymatic activity as a 5 , 6-dihydroxyindole-2-carboxylic acid ( DHICA ) oxidase that is active in melanin synthesis [25] . However , TYRP1 has also been reported to have a catalase function , and as such could also broadly influence cellular reactions to oxidative stress [26] . The findings of both the candidate-driven and genetic background-driven approaches suggested that Lyst influences oxidative stress associated with melanin synthesis . To independently test this hypothesis , lipid hydroperoxide and protein oxidation levels were measured from iris lysates of 2–3 month-old mice ( Figure 5 ) . All contexts of Lyst mutation resulted in significantly higher lipid hydroperoxide levels compared to strain-matched controls ( Figure 5A ) . Lyst genotype , genetic background , and the interaction between Lyst genotype and genetic background all significantly influenced lipid hydroperoxide levels ( P<0 . 001 in all comparisons , two-way ANOVA ) . In contrast , while genetic background significantly influenced protein carbonylation levels ( P = 0 . 003 , two-way ANOVA ) , Lyst genotype did not ( Figure 5B ) . Indices of accumulated oxidative lipid damage were also observed from immunohistochemical analysis of 4-HNE localization ( Figure S5 ) . Thus , the Lystbg-J mutation specifically altered the accumulation of oxidative damage in the membrane compartment . Importantly , the elevation in lipid hydroperoxide levels observed in the context of different genetic backgrounds mirrored the extent of disease sensitivity ( B6 . Tyrc-2J Lystbg-J<B6-Lystbg-J<D2 . Lystbg-J ) . This correlation is consistent with the previous finding that Lyst mutation impairs lysosomal exocytosis , which is important for plasma membrane repair [11] , and supports the notion that oxidative membrane damage contributes to the pathology of Lyst-mutant phenotypes . Oxidative membrane damage resulting from aberrant LYST function could have particularly important ramifications for the neurodegenerative component of Chediak-Higashi syndrome [1] , [27] . Elevated levels of oxidized lipids have been observed in several neurodegenerative diseases [28] . It is possible that the same process responsible for rapid degeneration of the iris might , over a longer time frame , contribute to damage in cells that are challenged by other forms of oxidative stress , for example in aging neurons . Supporting this , extensively aged D2 . Lystbg-J mice spontaneously developed a severe tremor indicative of a neurodegenerative phenotype , whereas B6-Lystbg-J mice did not ( Video S1; n = 5 mice per strain , 17–20 months in age ) . Further histologic analysis indicated that these D2 . Lystbg-J mice exhibited Purkinje-cell degeneration ( Figure 6 ) . Although the D2 . Lystbg-J mouse cerebellum was normal in overall size and lobule morphology ( Figure 6A and 6B ) , it consistently contained focal areas lacking Purkinje cells ( Figure 6C–6F; n = 5 mice per strain , 17–20 months in age ) . Analysis of sections from the spinal cord and sciatic nerve failed to show any degenerative pathology , suggesting limited , if any , lower motor neuron or peripheral nerve involvement ( Figure S6 ) . These findings indicate that the DBA/2J genetic background also uncovered a Lyst-mediated phenotype in the CNS , causing a tremor likely mediated by Purkinje-cell degeneration . Because of the requirement for extensive aging , it is not yet known whether the DBA/2J-derived modifier ( s ) of this neurodegenerative phenotype and the degenerative iris disease are identical . However , given that Tyrp1 is expressed in the brain ( Figure S7 ) and is thought to exhibit catalase activity [26] , this seems likely . In order to test whether the observed Purkinje-cell degeneration also involves oxidative damage to the cell membrane , lipid hydroperoxide levels were measured in cerebellar lysates of B6-Lystbg-J and D2 . Lystbg-J mice ( n = 4 mice per strain , 17–20 months in age ) . An average 25% elevation in lipid hydroperoxides was observed in cerebella of D2 . Lystbg-J mice compared to B6-Lystbg-J mice , but the trend was not statistically significant ( P = 0 . 20 , two-way ANOVA ) . Although no histologic defects were apparent in the cerebral cortex or brain stem ( data not shown ) , lipid hydroperoxides in the cortex were elevated by an average of 14% ( P<0 . 001 , two-way ANOVA ) , and levels in the brain stem by 51% ( P = 0 . 04 , two-way ANOVA ) . Despite the limited statistical power of these results , they suggest that , as in the iris , the sensitivity of Lyst-mutant neuronal phenotypes in the CNS may involve elevated lipid hydroperoxide levels .
Here we have extended knowledge of Lyst-mediated phenotypes through studies of Lyst genetic modifiers . Taking advantage of iris phenotypes as a convenient assay , two genetic contexts with important modifying influences were identified . Albinism completely rescued Lyst-mutant iris phenotypes , and the DBA/2J genetic background enhanced them . Both results implicate melanosomes in progression of disease associated with Lyst mutation . Because melanin production occurring in melanosomes is a potent source of reactive oxygen species , the iris of all three strains was tested for indices of oxidative stress by measuring levels of protein and lipid oxidation . These experiments demonstrated that in pigmented cells , Lyst mutation specifically results in oxidative damage to lipid membranes , which correlates with the overall phenotypic severity of iris phenotypes observed among the enhancer and suppressor strains ( B6 . Tyrc-2J Lystbg-J < B6-Lystbg-J < D2 . Lystbg-J ) . Thus , these experiments with Lyst genetic modifiers suggest that one mechanism contributing to Lyst-mutant phenotypes is oxidative membrane damage . B6-Lystbg-J mice have previously been described to exhibit multiple features of Chediak-Higashi syndrome [1] , as well as an iris disease recapitulating aspects of exfoliation syndrome [4] , [18] . In mice , both disease associations are characterized by changes to pigmented tissues , including coat color and iris morphology . From a mechanistic perspective , these results are directly relevant to the pathophysiology of Lyst-mutant defects in melanosomes . Eumelanin production occurring in melanosomes is known to be a potent source of oxidative stress [29] , [30] . The mechanisms that protect melanosomes and pigment-producing cells from this insult are not well understood . Our current findings support the hypothesis that Lyst influences these events by modulating the repair of oxidatively damaged membranes . Exocytosis of intracellular vesicles plays an important role in plasma membrane repair [31] , and experiments with cultured cells have previously demonstrated that Lyst mutations cause defects in lysosomal exocytosis and plasma membrane repair [11] . The oxidative membrane damage observed in the iris may well represent an accumulation caused by deficient repair . Thus , other defenses against oxidative damage to lipids are presumably overcome , leading to elevated levels of oxidatively damaged membranes and , ultimately , cellular demise [32]-[34] . The identification of Tyrp1 as a likely modifier of Lyst-mutant phenotypes challenges common notions of TYRP1 function . TYRP1 is typically ascribed to function as a melanocyte-specific protein involved in melanin synthesis with DHICA oxidase activity . However , human TYRP1 appears to lack DHICA oxidase activity [35] , indicating that this activity is not evolutionarily conserved . Furthermore , Tyrp1 is not exclusively expressed in only pigment producing cells where DHICA is found . Based on our results and data provided in online databases such as the Allen Institute for Brain Science's Mouse Brain Atlas [36] , Tyrp1 is also expressed in the brain . An alternative function for TYRP1 that is consistent with our current findings is to provide catalase activity [26] . A function for TYRP1 as a catalase that influences reactive oxygen species would be consistent with the observation that the Tyrp1b mutation is associated with elevated oxidative stress , and would provide a rational explanation for its ability to enhance Lyst-mediated oxidative membrane damage . However , in considering potential links between Tyrp1 and Lyst , it is important to point out a caveat of our current experiments . The D2-derived modifier has formally been mapped only to a congenic interval containing Tyrp1 . Given that the Tyrp1b and Lystbg-J mutations independently cause similar phenotypes in the iris , it is highly likely that Tyrp1 is the causative modifier , yet it remains possible that an as yet unknown modifier exists in close proximity to this gene . Experiments testing this directly are underway . Our current findings have important implications with respect to Chediak-Higashi syndrome . A defining component of this syndrome is progressive neurologic dysfunction [1] . Although bone-marrow transplantation can correct the immunological aspects of Chediak-Higashi syndrome and significantly extend lifespan , this treatment does not correct the neurologic aspects of the disease [37] , [38] . A deeper understanding of LYST-mediated neurodegenerative phenotypes is critical for the eventual development of improved therapies for this condition . In the current analysis , a change in genetic background has uncovered a neurodegenerative phenotype involving the loss of Purkinje cells in mice with the widely utilized Lystbg-J mutation . Our preliminary experiments suggest that , as in the case of the iris , the neuronal phenotype may involve an accumulation of oxidatively damaged membranes . Due to the large size of the neuronal cell and its expansive plasma membrane [39] , neurons are likely to be in need of continuous membrane repair , and especially sensitive to defects in this process . D2 . Lystbg-J mice represent a new resource for further dissecting these mechanisms , and for testing various anti-oxidant therapies for potential benefit in mouse models of Chediak-Higashi syndrome . Our findings also have important implications with respect to ophthalmic disease . The ocular phenotypes of B6-Lystbg-J mice , particularly the iris transillumination defects , resemble those seen in exfoliation syndrome [4] . Several studies implicate oxidative stress as contributory to exfoliation syndrome [5] , including the observation that aqueous humor from exfoliation syndrome patients has decreased levels of catalase activity [40] . The results presented here suggest that such changes are likely to be pathological . Furthermore , LYST , and other genes influencing oxidative stress , are suggested as candidates worthy of consideration for contributing to hereditary forms of exfoliation syndrome which is likely to also be strongly influenced by genetic modifiers [41] . Despite the existence of many Lyst alleles in mice , the resource that this allelic series represents has only begun to be utilized in assigning genotype-phenotype correlations . The bg-J mutation utilized here results from a 3-bp deletion predicted to remove a single isoleucine from the WD40 domain of the LYST protein [4] . Previous western blot analysis of cultured fibroblasts homozygous for the bg-J mutation failed to detect LYST protein [42] , suggesting that the mutation may represent a null allele . However , this experiment has not been performed on tissues isolated directly from the mouse , nor have genetic complementation tests with a definitive null ( such as a deletion or targeted mutation ) been performed , leaving uncertainty regarding classification of the bg-J allele . Neurodegenerative phenotypes have previously been described for only one other allele ( LystIng3618 ) [43] , which like the bg-J mutation , also disrupts the LYST WD40 domain . To our knowledge , iris phenotypes have not yet been assessed in any Lyst mutant strains other than those described here . Thus , it is not yet clear whether the iris and neuronal phenotypes described here will pertain to all Lyst alleles or might be specific to just a sub-class of mutations , though this is an issue that is addressable and worthy of follow-up . In addition to mutations in mice , a variety of mutations relevant to LYST have been identified in other model organisms . One example is the Drosophila BEACH family member , blue cheese ( bchs ) . Like LYST , the Bchs protein is predicted to be a large ( 400 kDa ) protein containing a BEACH domain followed by a series of WD40 repeats near the C-terminus . Unlike LYST , Bchs also contains a PI ( 3 ) P-binding FYVE domain . Mutations in bchs result in reduced adult life span and age-related neuronal degeneration [44] . The bchs gene exhibits genetic interactions with genes involved in lysosomal transport and is therefore thought to encode a scaffolding protein involved in vesicle transport [45] . In motor neurons from bchs mutants , anterograde transport of endolysosomal vesicles toward synaptic termini is particularly affected , leading to a hypothesis that a degradative function of endolysosomal compartments at the neuromuscular junction is important in preventing neuron degeneration [46] . With respect to our current findings with Lyst mutant mice , these observations demonstrate that lysosomes undoubtedly make several contributions important to neuronal survival and point to the opportunity afforded by experiments with model organisms to study these events . A direct Lyst ortholog exists in Drosophila ( CG11814 ) , but mutant phenotypes associated with this gene have not yet been described . In the future , it will be interesting to examine the extent to which bchs and CG11814 mutant phenotypes resemble each other and what additional insights might be gained by genetic studies of these genes . In conclusion , we have performed both candidate-driven and genetic background-driven experiments to identify Lyst modifiers . A priori , the expectation would have been that modifiers of Lyst would logically be related to organelle biogenesis . Instead , it seems that at the level of the whole animal , oxidative damage to membranes is a highly relevant event . In our ongoing work , we intend to further test the links between Tyrp1 and Lyst-mediated ophthalmic disease , and to dissect the neurodegenerative disease uncovered in D2 . Lystbg-J mice .
C57BL/6J , B6-Lystbg-J/J ( abbreviated throughout as B6-Lystbg-J ) , DBA/2J , and B6 ( Cg ) -Tyrc-2J/J ( abbreviated throughout as B6 . Tyrc-2J ) mice were obtained from The Jackson Laboratory , Bar Harbor , Maine . D2 . Tyrp1B6GpnmbB6 mice [24] were kindly provided by Dr . Simon John of The Jackson Laboratory and subsequently bred at The University of Iowa . Unless otherwise noted , all experiments with B6-Lystbg-J mice utilized mice homozygous for the bg-J mutation . All mice utilized were housed and bred at the University of Iowa Research Animal Facility . Mice were maintained on a 4% fat NIH 31 diet provided ad libitum and were housed in cages containing dry bedding ( Cellu-dri; Shepherd Specialty Papers , Kalamazoo , MI ) . The environment was kept at 21°C with a 12-h light:12-h dark cycle . All animals were treated in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research . All experimental protocols were approved by the Animal Care and Use Committee of The University of Iowa . Anterior chamber phenotypes were assayed using a slit-lamp ( SL-D7; Topcon , Tokyo , Japan ) and photodocumented using a digital camera ( D100; Nikon , Tokyo , Japan ) . All ocular exams utilized conscious mice . Based on previous observations of Lyst-mutant mice [4] , several traits uniformly present in adult B6-Lystbg-J mice were followed for potential phenotypic modification . For assessment of anterior chamber phenotypes , a beam of light was shone at an angle across the eye , and the anterior chamber was examined for iris stromal atrophy , pigment dispersion , and dark iris appearance . For assessment of iris transillumination defects , a small beam of light was shone directly through the undilated pupil of the mouse and the iris was examined for the ability of reflected light to pass through diseased or depigmented areas of the iris . All photographs of like kind were taken with identical camera settings and prepared with identical image software processing . Unless otherwise noted , all slit-lamp images were collected at 25× magnification , cropped , and reduced in size . Severity of iris transillumination defects was quantified by measuring the R-value from RGB formatted digital images . Digital images of iris transillumination defects from left and right eyes of 4 mice per genotype were analyzed using Adobe Photoshop software ( Adobe Systems Inc . , San Jose , CA ) . From 2 images per eye , 2 circular sampling windows of equivalent size , each covering approximately 5% of the measurable area of the iris , were uniformly placed ( 1 superior and 1 inferior ) on the temporal halves of each iris image using the Elliptical Marquee tool . RGB values for the sample areas were averaged using the Average Blur Filter and R-values measured with the Eyedropper tool . In total , each genotype of mice involved analysis of 32 sample areas whose R-values were utilized in statistical analysis . Samples from different tissues were processed as explained below , and imaged using a light microscope ( BX52; Olympus , Tokyo , Japan ) equipped with a digital camera ( DP72; Olympus , Tokyo , Japan ) . Eyes were fixed in 2 . 5% gluteraldehyde in 0 . 1 M Na cacodylate for 16 hours , and post fixed with 1% osmium tetroxide in 0 . 1 M Na cacodylate buffer at room temperature for 1 hour . A series of acetone dehydrations were performed followed by infiltration with Embed-812/DDSA/NMA/DMP-30 for 24 hours . 0 . 5-µm sections were cut ( EM UC6 ultramicrotome; Leica , Wetzler , Germany ) , and stained with 1% toluidine blue . Cerebella were cut down the midline , yielding 2 hemispheres . The left cerebellar halves were fixed overnight at 4°C in 4% paraformaldehyde in 1X PBS ( pH 7 . 4 ) , and embedded in paraffin ( Tissue Prep Paraffin Beads T565; Fisher , Pittsburgh , PA , USA ) . Mid-sagittal 5-µm sections were cut ( Microm HM 355; Thermo Fisher , Waltham , MA , USA ) and stained with hematoxylin-eosin ( H&E ) . Sciatic nerves were removed from the left hindlimb and fixed at 4°C in 2 . 5% osmotically-balanced glutaraldehyde in 0 . 1 M Na cacodylate buffer ( pH 7 . 4 ) for at least 24 hours . Following rinses with cacodylate buffer , nerves were post fixed with 1% osmium tetroxide in 0 . 1 M Na cacodylate buffer at room temperature for 1 hour . Dehydration was then carried out through a series of 40-minute incubations in 25% , 50% , 75% , 90% , and 100% graded ethanol . Nerves were infiltrated overnight at room temperature , with 33% , 66% , and 100% resin ( Low Viscosity Spurr Epoxy Resin; Ted Pella , Redding , CA ) in propylene oxide . Specimens were embedded in resin , and 1-µm cross sections were cut ( EM UC6; Leica , Wetzler , Germany ) and stained with toluidine blue . Dissected spinal columns were fixed in Bouins fixative for >1 week . Following rinses with 70% ethanol , 3-mm cross sections were cut from the cervical , thoracic , and lumbar regions of each column . The 3 cross sections from each column were embedded in paraffin ( Tissue Prep Paraffin Beads T565; Fisher , Pittsburgh , PA ) , and 5-µm cross sections were cut ( Microm HM 355; Thermo Fisher , Waltham , MA , USA ) and stained with H&E . The Lystbg-J mutation results from a 3-bp deletion predicted to remove a single isoleucine from the WD40 domain of the LYST protein [4] . Lyst genotype was assessed by PCR amplifying a fragment of genomic DNA that flanks the causative 3-bp bg-J deletion [4] and assessing product lengths . To generate B6 ( Cg ) -Tyrc-2J Lystbg-J mice ( abbreviated throughout as B6 . Tyrc-2J Lystbg-J ) , B6 . Tyrc-2J mice were bred to B6-Lystbg-J mice , and each region was bred to homozygosity . The Tyrc-2J allele is a spontaneously arising missense mutation that also influences splicing of the tyrosinase pre-mRNA , ultimately resulting in complete absence of the tyrosinase protein [47] . Tyr genotype was inferred from coat color . To generate congenic mice with the Lystbg-J mutation on a DBA/2J genetic background ( D2 . B6-Lystbg-J/Andm; abbreviated throughout as D2 . Lystbg-J ) , B6-Lystbg-J mice were reiteratively bred to DBA/2J mice and each successive generation genotyped to select breeders heterozygous for the Lystbg-J mutation . This process was continued for 10 generations of backcrossing . At the 10th generation , the mice were intercrossed and the Lystbg-J mutation was bred to homozygosity . Congenic mice were genotyped with the closely linked D13Mit17 marker and confirmed by genotyping of the causative 3 bp bg-J deletion . Genotypes of ( D2 . Lystbg-J X D2 . Tyrp1B6GpnmbB6 ) F2 progeny were assessed for Tyrp1 using the flanking markers D4Mit327 and D4Mit178 , and for Gpnmb using D6Mit355 and D6Mit74 . For reverse-transcription PCR ( RT-PCR ) , brains were removed and the cerebral cortex , cerebellum , and brain stem were dissected in PBS . Samples were homogenized and RNA was extracted , treated with DNase I , purified ( Aurum Total RNA Mini Kit; Bio-Rad Laboratories; Hercules , CA ) , and converted to cDNA ( iScript cDNA Synthesis Kit; Bio-Rad Laboratories; Hercules , CA ) . Each PCR reaction contained: 1 . 5 µl 10X reaction buffer ( Bioline , Taunton , MA ) , 1 . 2 µl dNTPs , 0 . 25 µl 5′-primer ( 10 µM ) , 0 . 25 µl 3′-primer ( 10 µM ) , 0 . 25 µl MgCl2 , 9 . 55 µl dH2O , 0 . 15 µl Taq DNA polymerase ( Immolase; Bioline , Taunton , MA ) , and 1 . 5 µl cDNA ( 0 . 66 ng/ul ) . Primer pairs used in PCR reactions include: Lyst ( 5′-CACTGGGAGCAAGTGTGGTG-3′ , 5′-TCAATTTCTGAGGGCGTGCT-3′ ) and Tyrp1 ( 5′-TGCGATGTCTGCACTGATGA-3′ , 5′-TCCAGCTGGGTTTCTCCTGA-3′ ) . PCR conditions were: 94°C for 10 minutes , 40× ( 94°C for 30 seconds , 61°C for 1 minute , 72°C for 1 minute ) , and 72°C for 7 minutes . PCR products were analyzed on a 1% agarose gel using EtBr detection . Lipid hydroperoxide levels were measured directly ( Lipid Hydroperoxide Assay; Cayman Chemical Company , Ann Arbor , MI ) following the manufacturer's protocol . To measure lipid hydroperoxide levels in the iris , both irides of individual mice were dissected and sonicated in ice-cold ddH2O . Total protein was measured from a small aliquot ( Bio-Rad Protein Assay; Bio-Rad Laboratories , Hercules , CA ) . The remaining sample volume was used for the extraction of lipid hydroperoxides into chloroform , and absorbance was measured following the protocol recommended by the manufacturer . Tissue lipid hydroperoxide was expressed as nmol hydroperoxide per mg of total protein . Each mouse strain was measured in biological triplicates from mice 2–4 months of age . Two-way analysis of variance ( ANOVA ) was employed to evaluate differences among different Lyst genotypes and genetic backgrounds . For measurement of lipid hydroperoxide levels in the brain , the right cerebellar and cortical hemispheres , as well as the right half of the brain stem , were dissected and processed as described above , with the exception that samples were analyzed in two separate sessions . Two mice of each genotype were analyzed per session . The percent difference was calculated by randomly pairing B6-Lystbg-J and D2 . Lystbg-J samples from each session . Two-way ANOVA was employed to evaluate differences among results from different genetic backgrounds and sessions . Protein oxidation levels were measured using an immunoblotting method ( OxyBlot™ Protein Oxidation Detection Kit; Millipore , Bedford , MA ) following the manufacturer's protocol . Dissected irides were homogenized in lysis buffer ( 50 mM Tris HCl pH 7 . 4 , 0 . 15 M NaCl , 1 mM EDTA , 0 . 1% TritonX100 , 0 . 1% SDS , and protease inhibitors ) . Two aliquots of 20 µg protein each were analyzed . One aliquot was subjected to the derivatization reaction and the other served as a negative control by substituting 1× derivatization-control solution for 1× DNPH solution . Samples were denatured , derivatized , and neutralized , followed by analysis of immunoblots using an αDNP antibody . Each sample was blotted in technical triplicates ( 4 µl per dot ) on 2 separate membranes ( Immobilon-FL PVDF; Millipore , Bedford , MA ) and allowed to dry completely . One membrane was stained with coomassie blue and the other was blocked ( Odyssey Blocking Buffer; LI-COR Odyssey , Lincoln , NE ) and incubated with αDNP primary antibody ( diluted 1∶50 in blocking buffer overnight at 4°C ) . Following washes and incubation with secondary antibody ( IRDye 680 Conjugated Gt α Rb IgG; LI-COR Odyssey , Lincoln , NE ) , blots were washed and quantified ( LI-COR Odyssey detection system; LI-COR Odyssey , Lincoln , NE ) . Each sample was normalized to total protein and the amount of protein oxidation in the wild-type C57BL/6J strain . Sample analysis was repeated in 5 independent experiments . Normalized values were averaged and compared by two-way ANOVA . For immunohistochemistry with 4-HNE , eyes from 6 month old mice were embedded unfixed in Optimal Cutting Temperature embedding medium ( Tissue-Tek O . C . T . Compound; Sakura Finetek U . S . A . , Inc . , Torrance , CA ) ; 10-µm sections were cut and sections were transferred to glass slides ( CryoJane , Instrumedics , Inc . , St . Louis , MO ) . Cryosections were air dried for 30 minutes at room temperature , fixed for 5 minutes in ice-cold acetone , again air dried for 30 minutes at room temperature , and rehydrated in PBS for 5 minutes . Sections were blocked ( 15 minutes at room temperature with 1 mg/ml BSA in PBS ) , labeled with primary antibodies ( 1 hour at room temperature using monoclonal mouse Anti-4-HNE antibody diluted 1∶100; Oxis International Inc . , Foster City , CA ) , washed ( three washes , 5 minutes each in PBS ) , and labeled with secondary antibody ( 1 hour at room temperature using AlexaFluor 488 conjugated antibody diluted 1∶200; Invitrogen-Molecular Probes , Carlsbad , CA ) . After 3 washes in PBS , the sections were mounted ( ProLong Gold , Invitrogen-Molecular Probes , Carlsbad , CA ) , and viewed by fluorescence microscopy . All immunohistochemical experiments used assay conditions in which controls using no primary antibody lacked specific signal . | LYST is a poorly understood protein involved in hereditary disease . Mutations in the encoding gene cause Chediak-Higashi syndrome , a rare lethal disease affecting multiple tissues of the body . Mutations in Lyst also recapitulate features of exfoliation syndrome , a common disease affecting the anterior chamber of the eye . Unfortunately , the Lyst gene is quite large , rendering it difficult to study by many molecular and cellular approaches . Here , we use a genetic approach in mice to identify additional genetic pathways which might modify , or prevent , the ill consequences associated with Lyst mutation . Our experiments demonstrate that Lyst mutation results in elevated levels of oxidative damage to lipid membranes . These results identify a previously unrecognized consequence of Lyst mutation and a modifiable pathway of potential clinical relevance in humans . Ultimately , knowledge of these events will contribute to the design of new therapeutic strategies allowing a similar alleviation of disease in humans . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/animal",
"genetics",
"cell",
"biology/membranes",
"and",
"sorting",
"genetics",
"and",
"genomics/disease",
"models",
"ophthalmology/inherited",
"eye",
"disorders",
"ophthalmology/glaucoma"
] | 2010 | Elevated Oxidative Membrane Damage Associated with Genetic Modifiers of Lyst-Mutant Phenotypes |
Conventional wisdom holds that the best way to treat infection with antibiotics is to ‘hit early and hit hard’ . A favoured strategy is to deploy two antibiotics that produce a stronger effect in combination than if either drug were used alone . But are such synergistic combinations necessarily optimal ? We combine mathematical modelling , evolution experiments , whole genome sequencing and genetic manipulation of a resistance mechanism to demonstrate that deploying synergistic antibiotics can , in practice , be the worst strategy if bacterial clearance is not achieved after the first treatment phase . As treatment proceeds , it is only to be expected that the strength of antibiotic synergy will diminish as the frequency of drug-resistant bacteria increases . Indeed , antibiotic efficacy decays exponentially in our five-day evolution experiments . However , as the theory of competitive release predicts , drug-resistant bacteria replicate fastest when their drug-susceptible competitors are eliminated by overly-aggressive treatment . Here , synergy exerts such strong selection for resistance that an antagonism consistently emerges by day 1 and the initially most aggressive treatment produces the greatest bacterial load , a fortiori greater than if just one drug were given . Whole genome sequencing reveals that such rapid evolution is the result of the amplification of a genomic region containing four drug-resistance mechanisms , including the acrAB efflux operon . When this operon is deleted in genetically manipulated mutants and the evolution experiment repeated , antagonism fails to emerge in five days and antibiotic synergy is maintained for longer . We therefore conclude that unless super-inhibitory doses are achieved and maintained until the pathogen is successfully cleared , synergistic antibiotics can have the opposite effect to that intended by helping to increase pathogen load where , and when , the drugs are found at sub-inhibitory concentrations .
Our arsenal of antimicrobials boasts a wide diversity of drugs and we continue to invest in the search for new ones [1] . Yet bacteria adapt so readily to their ambient environment that all antibiotics in clinical use have bacteria that resist them [2] , [3] . A Staphylococcus aureus infection traced in vivo yielded over thirty de novo mutations from a 12-week therapy , each mutation conferring an increase in drug resistance [4] . With such a rapidly evolving foe and antibiotic discovery programmes waning substantially [3] , determining optimisation principles that maintain the efficacy of the antibiotic repertoire already in our possession represents one of the keenest challenges confronting the scientific community . And yet drug-resistance evolution has been called ‘conceptually uninteresting’ [5] . This view is the result of assuming a fixed timeline: a pathogen is treated with antibiotics , resistance traits emerge , sweep through the population and fix . The more efficient the drug , the greater selection for resistance and the sooner resistance fixes . The only mitigating action we can take is hit early , hit hard and kill drug-susceptible cells before they accumulate , so the old argument goes [6] . Bacteria are hardest hit by multi-drug combinations . Developed for over 70 years [1] , [7] , [8] , combinations are key in our fight against microbes [9] , viruses [10] and cancers [11] . Combinations said to be synergistic , where two drugs hit the pathogen much harder than each drug alone , are highly prized [1] , [12] , [13] . Indeed , the rapid deployment of synergistic antibiotics should , according to the same logic , be the fastest way of clearing a bacterium . To make our discussion more precise we say that a pair of bacteriostatic antibiotics of equal efficacy is synergistic if a 50-50 weighted combination of both drugs inhibits growth more than the two single-drug treatments when measured over one day of bacterial growth [8] , [14]–[16] . ( Strictly speaking , we ask this for all ( θ , ( 1−θ ) ) -combinations where θ is any value between 0 and 100% , not just 50-50 , as shown in Figure 1 . ) With this definition we can formulate a null hypothesis , H0: a synergistic drug combination also inhibits growth synergistically if the treatment lasts longer than a day . Put differently , if the 50-50 combination treatment is more efficient than both single-drug monotherapies on the first day of treatment , it should also be more efficient on subsequent days to be deemed synergistic . Any in vitro test of H0 necessitates the use of antibiotic concentrations that support measurable population densities , the treatments we can use to test it are , as a result , necessarily constrained to a sub-inhibitory dosing regime . We must therefore question how relevant this study can be to antibiotic use in vivo , we argue that it is relevant for the following reasons . Drug interactions are often determined by one-day checkerboards and isoboles [17] , like those illustrated in Figure 1 , but by their very nature checkerboards only provide insight into the interaction inside the sub-inhibitory regime as isoboles can only be calculated if cells grow . Moreover , drug concentrations can sweep downwards from their highest values to sub-inhibitory concentrations during treatment ( [18] , Figure 1 ) , repeatedly so for intermittent dosing regimens [19] , [20] . The different diffusivities small antibiotic molecules exhibit in different tissue can create substantial inhomogeneities in concentration [21] resulting in a potential spatiotemporal mosaic of selection for resistance [18] , [22] whereby treatment can reduce pathogen load in some , but not all , organs [23] . Indeed , spatial diffusion itself creates concentration gradients with rapid , super-exponential decay away from a point source . It is therefore essential to understand how antibiotic combinations mediate resistance at all dosages within this mosaic , including sub-inhibitory , particularly as resistance is known to be selected for at very low concentrations , well below the minimal inhibitory concentration [24] . Now , we argue that treatments with the greatest short-term efficacy do not necessarily lead to the lowest bacterial densities later . A simple construction accounting for both density-dependent and frequency-dependent selection on drug resistance suffices to explain why . Consider three scenarios with two drugs , ‘A’ and ‘B’ . A bacterium is either unchallenged by antibiotics , challenged with drug A only ( or drug B only ) or else treated with the optimally synergistic combination of both , as in Figure 2 ( a ) . The no-drug treatment sees the cells grow , to carrying capacity say , without selecting for drug-resistant phenotypes . The synergistic combination inhibits drug-susceptible cells optimally , better than the two monotherapies , and so , by the end of day 1 , the lowest bacterial load of all is observed in this treatment . However , suppose some cells exhibit genetic or epigenetic adaptation conferring resistance; such cells may even have been present in low frequencies at the start of treatment . It is now in the synergistic line that drug-resistant phenotypes fare best as they have fewer competitors for the extracellular metabolites needed for growth . To clarify how this might arise , imagine a population of bacteria with two subpopulations of drug-susceptible and resistant cells and suppose extracellular metabolites are shared equally among all the growing cells . As the growth of susceptibles is suppressed more at greater synergies , more metabolites become available for resistant cells in those treatments . However , resistant cells necessarily grow faster than susceptible cells do when the drugs are present , with a greater fitness difference at greater synergies . Thus the total population density can be increased by the synergy even when the number of drug-susceptible cells present is reduced . Now , if resistant cells are absent or at low frequencies at the beginning of treatment , the exposure to antibiotics must be long enough to allow the resistants to achieve densities comparable to the susceptibles and so the treatment duration then needs to be long enough for the claim in the previous sentence to be true . This process is illustrated in Figure 2 . This idea , known as ‘competitive release’ [25] has been tested in treatments of malaria in vivo using mice [5] where higher drug concentrations have been shown to select for higher parasite load but competitive release makes new predictions for antibiotic therapy , for combinations in particular . First , the optimal combination is not robust: the best way of deploying a drug pair depends on how long the treatment lasts . Second , and as a result , the favoured property of antibiotic synergy is not necessarily robust to adaptations that confer drug resistance . Not only will synergy decay with time , it can be lost completely and replaced with an antagonism because more potent combinations have paradoxically selected for larger bacterial load . Thus the theory of competitive release is not consistent with our null hypothesis and provides an evolutionary rationale for rejecting it . A toy mathematical model captures the verbal argument completely and shows that synergy loss can be viewed as a form of tipping point . Imagine a bacterial population consisting of cells susceptible to both antibiotics at density S ( t ) , where t is time . Suppose there is a completely resistant phenotype , R ( t ) , and μ is the mean rate in a random Poisson process by which susceptible cells gain resistance . The dimensionless variable θ between zero and one controls the drug combination and k ( θ ) = 1+θ ( 1−θ ) measures the efficiency of each combination at drug concentrations ( A , B ) = ( A0θ , B0 ( 1−θ ) ) . Here A0 and B0 are normalising concentrations , chosen so that each drug achieves equal inhibitory effect at a defined time . Note that k ( θ ) is maximised when θ = 1/2 . This value represents a 50-50 combination therapy whereby ( A , B ) = ( A0/2 , B0/2 ) . The toy model is the following logistic growth equation modified to include antibiotics: ( 1a ) ( 1b ) where and R ( 0 ) = 0 . We therefore begin with susceptible cells but no resistant ones . Figure 2 ( b ) shows the population densities that result from this model , Δt ( θ ) = S ( θ , t ) +R ( θ , t ) , plotted as a function of θ for increasing values of time t . For short times ( Equation 1a–b ) exhibits synergy because density is suppressed most by the combination where θ = 1/2 , so the plot of Δt ( θ ) has the convex , U-shaped ‘smile’ shown in blue in Figure 2 ( b ) . At later times , but only provided μ>0 , the shape of the density profile changes and now density is greatest for the 50-50 combination and lowest for the ‘monotherapies’ , where θ = 0 and θ = 1 . So the plot of Δt ( θ ) now exhibits a near-concave , W-shaped ‘frown’ consistent with antagonism having its maximal value at θ = 1/2 , as shown in red in Figure 2 ( b ) . Density is now maximised where before it was minimised . We call the resulting passage from synergy to antagonism the ‘smile-frown transition’ , referring to it on occasion as ‘synergy inversion’ because the convex , synergistic profile is inverted to form a near-concave , antagonistic one; this is a different notion of synergy inversion to the one in [26] . If we set μ = 0 , thus preventing the modelled population from adapting to the drug , it then follows that Δt ( θ ) has a synergistic profile at all times . In this case the 50-50 combination , represented by the value θ = 1/2 , is the optimal combination for all times as it minimises population density , irrespective of treatment duration . We tested the veracity of these theoretical predictions using an evolutionary functional genomics approach that combined evolution experiments using Escherichia coli , a genomic analysis , the genetic manipulation of an identified candidate resistance mechanism and quantitative mathematical modelling . This approach highlights the molecular mechanism that causes the synergy loss predicted by theory , whereas the theory alludes to the generality of the empirical results that we now describe .
The above predictions are best tested in vitro where the drug interactions are well-understood and can be rigorously controlled . We therefore cultured E . coli K12 ( MC4100 ) over a five-day period using a serial dilution protocol and sixteen different combination treatments of erythromycin ( ERY , a macrolide ) and doxycycline ( DOX , a tetracycline ) , two bacteriostatic translational inhibitors with an established synergy [14] . The bacteria are first cultured for 24 h in liquid growth medium containing antibiotics at concentrations described below and , at the end of the 24 h period , a random sample of the bacteria is transferred using a standard plate replicator to inoculate fresh growth medium . This process is repeated to create a treatment lasting several days . We began by choosing a pair of normalising , or ‘basal’ , antibiotic concentrations , D50 and E50 , in such a way that each DOX-only and ERY-only monotherapy achieved a 50% reduction in density when measured at 24 h relative to a zero-drug control ( the basal concentrations D50 and E50 are the IC50 values of each drug ) . Each of the sixteen different treatments may therefore be described by a single pair of concentrations ( 2 ) where θ is the relative drug proportion . When combined in a 50-50 ratio at these doses , where θ = 1/2 , a 90% reduction in bacterial growth at 24 h is achieved , greater than the 50% reduction achieved by each monotherapy ( the data in Figure 3 ( a ) ( Day 1 ) supports this ) . We implemented 14 different combination treatments and two monotherapies at those basal doses with θ ranging in discrete values from 0 and 1/15 to 14/15 and then 1 ( 19 replicates per treatment; see Section 3 . 2 in Text S1 ) . The fixed drug proportion , θ , that minimises bacterial density from the sixteen implemented and determined empirically by culturing the bacteria for 24 h will be denoted by θsyn in the following . This value between zero and one denotes the maximally synergistic combination treatment obtained after fixing the basal drug concentrations , as shown in Figure 1 ( b ) . The time-dependent optimal combination will be denoted θopt ( T ) ( see Materials and Methods ) and this value represents the combination of ERY and DOX that minimises density for a treatment of duration T hours . It follows by design that θopt ( T ) = θsyn if T is small , less than 24 h , say . After calibrating concentrations D and E so that each drug has equal effect , so θsyn≈1/2 in practise as Figure 4 ( c ) shows , the short-term optimal treatment is a 50-50 combination of both ERY and DOX . As a reflection of this , the day 1 data in Figure 3 ( a ) then shows the 50% growth reduction obtained for each monotherapy , the 90% reduction for the maximally synergistic 50-50 combination in addition to the growth reduction for all the other combinations we tested . We can now test our null hypothesis by asking whether the drug combination that is optimal on day 1 , 50-50 by design , is also optimal on subsequent days . Equation 1 makes a clear prediction: the best therapy on day 1 will be the worst later . The first day's data exhibits synergism with the lowest short-term bacterial densities found for near 50-50 combinations of ERY and DOX , so θsyn≈1/2 , this can be seen in Figure 5 ( shown in blue ) . However , the subsequent population dynamics beyond day 1 lead to us to reject H0 for Figure 5 ( in red ) shows they are consistent with the theory of competitive release and exhibit the smile-frown transition before 36 h have elapsed , as we now explain . Consistent with the predictions of Equation 1 , Figure 4 ( a ) illustrates how the degree of interaction , I ( T ) , defined in Materials and Methods , shifts from synergy ( where I ( T ) <0; t-test , df = 19 , t≈−6 . 13 , p<0 . 0001 , ) to antagonism ( where I ( T ) >0; t-test , df = 19 , t≈6 . 83 , p<0 . 0001 ) between 12 h and 36 h . The degree of interaction thereafter remains positive , denoting antagonism , until the end of the experiment . This is shown with more detail in Figure 4 ( b ) where the dynamics of the interaction profile are shown on an hour-by-hour basis; this illustrates that the interaction changes at about 30 h . Examining the apparent change in drug interaction more closely in Figure 5 , at 12 h the interaction profile is synergistic ( α-test , α = 0 . 61±0 . 05>0 , df = 13 , t≈11 . 22 , p<10−7 , θopt ( 12h ) = 0 . 49±0 . 01≈θsyn; see Materials and Methods and Section 4 . 3 in Text S1 for a description of the α-test ) but combination treatments for which θ≈2/3 ( estimated robustly using the α-test described in Materials and Methods as 0 . 65±0 . 04 ) yield the highest observed population densities by 36 h . As a result , the optimal combination has changed within two days from a 50-50 combination to an ERY monotherapy because the interaction profile is now antagonistic ( α-test , α = −0 . 44±0 . 14<0 , df = 13 , t≈−3 . 05 , p<0 . 0093 , θopt ( 36h ) ≈0≠θsyn; Section 4 . 3 in Text S1 ) . These data were produced for optical density measures of bacterial growth , but analogous results are obtained using different notions of fitness . Using an area under the curve measure of growth inhibition that accounts for both population sizes and growth rates ( Section 4 . 2 in Text S1 ) Figure 3 ( a ) shows drug efficacy approaches zero most rapidly for near 50-50 combination treatments . The same figure shows the optimal treatment has shifted in this measure too , to the ERY monotherapy within two days . For completeness , the smile-frown transition is also seen if we use colony-forming units to measure bacterial population densities ( Section 7 . 4 in Text S1 ) . As a further test for loss of synergy , dose-response checkerboards and isobolograms were produced using bacteria sampled from the highly synergistic θ≈8/15 treatment at the beginning of days one and five , both are shown in Figure 6 . The earlier checkerboard is consistent with synergism whereas the latter checkerboard shows a progressing wave of increased resistance , with synergy at higher drug concentrations and a mixed interaction apparent at lower concentrations . Figure 6 ( a , right ) shows isoboles at 60% inhibition that are suggestive of a suppressive interaction by day 4 in which doxycycline reduces the inhibitory effect of erythromycin . The white isobole of 50% inhibition in Figure 6 ( b ) shows a shift from day 1 to 5 that indicates increased resistance of the population to both ERY and DOX ( for controls that the antibiotics do not degrade significantly when stored at 4°C for several days see Section 3 . 2 in Text S1 ) . Having established the rapid loss of optimality of the most synergistic combination treatment , at which point the latter becomes the worst treatment of all , it is essential we understand the genetic basis of this change . So we first performed a test to determine whether increased drug resistance was the result of epigenetic adaptation ( Section 3 . 2 in Text S1 ) . Samples each of the initially most synergistic drug treatment and the control treatment without drug were taken from the end of day 5 and cultured without antibiotics for a further 24 h . The resulting populations were then all subjected to the most synergistic drug combination for another 24 h . Consistent with a likely genetic basis to drug-resistance adaptation , samples from the short-term synergistic treatment still displayed greater AUC inhibition when measured relative to the no-drug control ( Wilcoxon signed rank test , W = 92 , N = 10 , p<0 . 001 ) . Knowing such rapid adaptation has a genetic basis , our goal was to exploit the resistance mechanism and understand what organismal function , if suitably manipulated , could maintain antibiotic synergy for longer and so ensure the smile-frown transition does not occur so rapidly . We therefore conducted a whole-genome sequencing study of independent biological replicates of both monotherapies and of the maximally synergistic treatment sampled at the end of day 5 . The analysis revealed single nucleotide polymorphisms ( SNPs ) in most replicates modifying physiology , metabolism and drug resistance , including treatments with SNPs in marRAB and acrR ( see Table 1 , Figure 7 , and Section 5 . 3 in Text S1 ) . Indeed , the mar regulon is known to control a range of stress-responses in E . coli [27] including the multidrug efflux system acrAB-tolC [28] . Rapid increases in resistance to antibiotics can occur when regions of the genome containing resistance genes are duplicated and whole-genome sequencing was proposed as a method to detect such duplications [29] , [30] . Our analysis revealed 90% of the independent replicates in the most synergistic combination treatment had the same 315 Kb fragment duplicated , a region containing several efflux pumps including acr ( Table 2 , Section 5 . 4 in Text S1 ) . The duplication was found in monotherapies too , but only in 30–40% of those treatments ( 3/10 replicates for DOX-only and 2/6 for ERY-only ) . The duplication was therefore observed significantly more for the 50-50 combination treatment than in the ERY monotherapy ( Fisher's exact test , P<0 . 035 ) and the DOX monotherapy ( Fisher's exact test , P<0 . 02 ) . In all 14 replicates where a duplication was detected , it was located between positions 274 , 201 bp and 589 , 900 bp . This region contains 293 genes , among which are 12 antibiotic resistance or binding genes , 32 transporter genes and 31 transposon-related genes ( Appendix B in Text S1 ) . Cross-resistance to antibiotics not used in the protocol is likely as three known multi-drug efflux systems and ampicillin degradation proteins are encoded within the duplicated region ( Section 5 . 4 in Text S1 and Appendix B in Text S1 ) . Such consistent , parallel evolution towards a 315 Kb duplication in all but one replicate of the 50-50 combination treatment strongly suggests , therefore , that genetic amplification of a multi-drug efflux pump is the adaptation that confers the multi-drug resistance phenotype we observe . To test the stronger hypothesis that a drug efflux system could be responsible for synergy loss and the smile-frown transition , we first developed a system-specific , physico-genetics theoretical model ( detailed in Section 6 . 4 of Text S1 ) in which cells may express a gene whose product can pump both antibiotics from the cell with no fitness or ATP cost . We assume the drugs have different affinities for the pump and the model encodes three phenotypes: drug-sensitive cells that do not express the efflux system , less sensitive cells that do and a third phenotype then possesses an additional efflux gene and expresses both . Figure 5 shows that the model successfully captures the first 48 h of data predicting that the rapid inversion of synergy that we observe empirically is consistent with the up-regulation and duplication of efflux genes . Generalising this mathematical framework , we can show that the short-term optimal combination , represented by θsyn , and the time-dependent optimal combination θopt ( T ) are close in general for a time that depends on the convexity of the drug interaction profile ( Section 8 . 2 in Text S1 ) . The two quantities are related as follows: ( 3 ) where T is treatment duration , ρ is the divergence rate between the optimal treatment and maximal synergy; ρ may be positive or negative depending on how the bacteria adapt to each drug . The times ( 4 ) are therefore approximations of the moment at which the optimal protocol is a monotherapy and no longer a combination . Figures 2 ( b ) , 3 and 5 all exhibit this phenomenon , but it can be seen most clearly in Figure 4 ( c ) that shows the dynamical path taken by the best and the worst therapies . Analogous to a critical transition , a shift takes place at 30 h of treatment where the 50-50 therapy displaces the DOX monotherapy as the worst treatment . The synergistic treatment never recovers its previously favourable status rather , as Figure 3 ( b ) shows in red , its performance continues to deteriorate exponentially . The physico-genetics model predicts the drug interaction profile will be robust to changes in the duration of treatment , which can be interpreted as ρ being reduced in magnitude and so synergy maintained , if the efflux system were suppressed ( Figure S16 in Text S1 ) . This is analogous to setting μ = 0 in Equation 1 above . To test this prediction we repeated the original evolutionary protocol using two new E . coli strains: a wild-type strain AG100 and a mutant AG100A ( Δacr ) [31]; we refer to Section 7 of Text S1 that details the minor differences between the first and now this evolutionary protocol . The latter strain differs from the former through a large deletion in acrAB that renders efflux systems that use the products of this operon , like acrAB-tolC , inoperable . As already observed using the E . coli K12 strain MC4100 , AG100 soon exhibited the smile-frown transition , within 48 h according to Figure 8 ( a ) . In contrast , the mutant strain AG100A ( Δacr ) that lacks acrAB continued to exhibit synergy until 72 h according to Figure 8 ( b ) , consistent with the prediction . We now ask whether the synergy loss we observe is contingent on the choice of D50 and E50 as basal drug concentrations . For example , might synergy be maintained for longer if we were to increase the dosage of both drugs ? We address this question with the following experiment . We re-ran the drug-specific mathematical model ( Section 6 . 4 in Text S1 ) at different dosages and repeated the evolutionary protocol using four different pairs of basal drug concentrations , chosen as follows . By analogy with ( 3 ) each new treatment can be represented by a pair of concentrationsEmpirically , we calibrated these four concentration pairs to produce a 40% , 80% , 90% and 95% reduction in growth relative to a zero-drug control by 18 h on day 1 for the 50-50 treatments ( ones with θ = 1/2 ) . We then subjected AG100 to treatments at each of the four basal dosages for a duration of five days using the drug proportions θ = 0 , 1/4 , 2/4 , 3/4 , and 1 . The prior mathematical model made a quantitative prediction for this new protocol that is depicted in Figure 9 ( a ) : the greater the antibiotic dose , the greater the synergy observed on day 1 and the greater the resulting antagonism on day 2 ( see also Figure 9 ( b ) ) . These figures show the model predicts that synergy is maintained from the first day onwards only when the dosages are sufficiently low . Figure 9 ( c ) shows the results of this experiment are in quantitative agreement with the model . Indeed , the numerical values of day-one synergy and day-two antagonism are positively correlated in both the model and the resulting data ( R2 = 0 . 990 , F = 145 , p<0 . 0069 ) provided the antibiotic dose is sufficiently high in the former . Finally , we observe more rapid selection for resistance at higher doses in the sense that the greater the dose , the sooner the transition to antagonism ( Section 7 . 3 in Text S1 ) .
It is important to state that we , of course , exercise extreme caution when drawing parallels with in vivo infections where the immune response , the highly-organised spatial structure of the host , xenobiotic metabolism and the pharmacokinetics that result may substantially complicate antibiotic interaction dynamics . However , we also argue that in vitro evolutionary studies of bacteria allied to genome-wide analyses and mathematical modelling can play an important role in elucidating how antibiotic interactions change through time precisely because model systems like ours are so simple . Drug interactions are subtle and synergy can be lost , and inverted , for reasons other than competitive release . Synergy must decay with time because of selection for drug-resistant alleles but it can be inverted when drugs degrade to produce non-antibiotic metabolites [26] . It is known that drug interactions can depend upon population heterogeneities because of differential pump expression between subpopulations [32] , but cellular mechanisms not commonly associated with resistance might also force drug interactions to change with time . For example , a theoretical model was used to propose [33] that synergism and antagonism could be found simultaneously in a population of cancer cells due to metabolic adaptation in subpopulations , the so-called Harvey Effect [34] . To our knowledge , this theory has not been tested . There are parallels with a prior study [14] that used antagonistic and synergistic antibiotic pairs to show that synergistic environments promote resistance more quickly than do antagonistic ones and the analogy of their result in our data is Figure 10 . Their core argument , that single drug-resistance mutations have a greater fitness effect in more synergistic environments is applicable to our study and consistent with our findings . Unlike ours , however , that study did not address which treatments lead to the lowest or greatest bacterial loads . Nothing of the molecular , multi-drug resistance mechanism is encoded within Equation 1 and despite its simplicity , this model may explain other phenomena . This includes the unreliability of antibiotic synergy assays such as checkerboards [35]–[37] . If a drug interaction assay were conducted with resistant cells in the inoculum [32] or if one emerged , irrespective of genetic mechanism , Equation 1 predicts synergy and antagonism could be reported for two replicates of the same checkerboard [37] . Indeed Figure 4 ( c ) illustrates how the change from synergistic to antagonistic interaction can occur quickly and it is only when population density data is sufficiently well-resolved through time that a transition point from one to the other is found . Our theoretical models are consistent with the smile-frown transition not being specific either to the drugs used or to the bacterium , any multi-drug resistance mechanism inactive in the absence of drugs that confers a fitness advantage in their presence may be sufficient ( Section 8 in Text S1 ) . However , while our data establishes that the duplication of a chromosomal multi-drug efflux operon is sufficient to observe the transition , this has been done for one Gram negative bacterial species and one drug pair . Many questions therefore remain regarding the generality of our observations . Clinically-important pathogens are known to efflux drugs into extracellular space , or the periplasm , thus conferring resistance to a wide range of drugs in many species [38] , [39] . As efflux has been observed both in clinical Staphylococcus aureus [40] and Mycobacterium tuberculosis ( TB ) [41] we ask whether synergy loss or the smile-frown transition might be observed in other bacteria . Relevant to this question is the study [35] of several clinical isolates of methicillin-susceptible and -resistant S . aureus ( MSSA and MRSA ) in which a combination of vancomycin and rifampin was variously reported as synergistic and antagonistic at 24 h and 48 h , with different interactions reported for both different strains and different drug concentrations . No mechanistic explanation has been attributed to this discrepancy and while this may not be at all related to efflux , the true nature of this important combination remains unclear [42] . What of drug combinations reliant on different mechanisms of synergy [43] ? The duplicated genomic region illustrated in Figure 7 contains dacA with β-lactamase activity [44] and three efflux systems in addition to acr . Efflux of fosmidomycin by far [45] , of aminoglycosides by emrE and of fluoroquinolones by mdlAB [39] , all of which are found in the duplicated region ( Table 2 ) , indicates the smile-frown transition may also be relevant to other classes of antibiotics . And would the transition still be observed if two target-altering , de novo mutations were needed for multi-drug resistance because there were no pre-existing chromosomal resistance mechanism that could be so rapidly duplicated ? We have not been able to determine a pair of such mutations and so , by way of a partial response , we compared the duration for which synergy is maintained when an important chromosomally-encoded multi-drug pump is , and is not , present using data from E . coli strains AG100 and AG100A ( Δacr ) . Figure 8 ( a ) shows that synergy is lost to antagonism in the former strain around 35 h but for the latter strain , the interaction only ceases to be significantly synergistic around 72 h , although significant antagonism is not observed thereafter . The latter strain , without acr , does therefore exhibit synergy loss but the smile-frown transition was not observed . However , in this case the interaction converges towards indifference in which one of the combination treatments maximises population densities by day 4 but without the smile-frown transition ever appearing ( Section 7 . 2 in Text S1 ) . It has been suggested that the treatment of multi-drug resistant TB will be more successful if supplemented with efflux pump inhibitors ( EPIs ) [39] , [46] . The present work suggests that if EPIs are used as an adjuvant to combination therapy they may prove beneficial by maintaining synergy for longer , although we have not conducted a direct test of this hypothesis using an EPI molecule . We conclude that complementary theoretical and in vitro approaches agree that the optimal way of combining antibiotics depends on the duration of treatment . This could have been deduced from a simple engineering principle that complex adaptive systems cannot be controlled optimally using strategies that are constant through time ( Section 8 . 2 in Text S1 ) . The consequences of this principle for antibiotic combinations are dramatic and cause the emergence of what looks like antagonism from a synergism , rendering the supposed optimal combination the worst treatment of all within a day . So while it is axiomatic in theory [18] and demonstrable empirically [14] that drug resistance rises faster for more synergistic treatments , that the greatest antibiotic potency can also select for the highest bacterial densities has been overlooked .
The protocol is a standard batch-transfer protocol used elsewhere [14] in the context of antibiotic treatments and described in detail in Section 3 of Text S1 . Briefly: bacteria are cultured in liquid growth medium for 24 h in the presence and absence of different antibiotics and continually shaken . Optical density measurements are taken continually from where the inhibition due to treatment can be calculated relative to the growth observed in a control cultured without drugs . After each 24 h period has elapsed , the environment is sampled and approximately 1% of biomass transferred to fresh a environment that includes replenished growth medium and drugs . This process was repeated for 5 days . There are many nonequivalent definitions of antibiotic synergy [8] , [17] , [47]–[49] . To ensure a precise quantification of drug interactions we use several consistent measures with different granularity derived with Loewe additivity as the key assumption . Suppose bacterial growth is measured over a fixed and short length of time , usually 24 h in the literature , although our measurements will be substantially longer . Population density is denoted by the function B ( D , E ) where D and E are extra-cellular drug concentrations , the number B ( 0 , 0 ) then represents density in a zero-drug environment . Assume each basal concentration , D and E , have been normalised to equal inhibitory effect , thus B ( D , 0 ) = B ( 0 , E ) = rB ( 0 , 0 ) . The value corresponds to the choice of IC50 for D and E , the concentrations denoted D50 and E50 in the text . Quantification of the drug interaction begins with i , the interaction profile , where i ( θ ) = B ( θD , ( 1−θ ) E ) . Following Loewe additivity [8] , i is said to be synergistic if , for all θ between zero and one exclusive , the effect of the drugs combined is greater than the sum of effects produced by each drug separately: ( 5 ) This definition is described pictorially in Figure 1 , Figures 1 ( b ) and 1 ( d ) are particularly relevant . Property ( 5 ) holds necessarily if i ( θ ) is convex ( c . f . blue lines in Figures 2 ( b ) and 5 ) . When property ( 5 ) does hold it follows that θsyn , the maximally synergistic drug proportion that satisfiesalso satisfies 0<θsyn<1 . Drug antagonism is said to occur when the reverse inequality applies in ( 5 ) , this is necessarily the case if i ( θ ) is concave . The drug interaction is additive in this context if i ( θ ) is independent of θ . Bacterial density is measured empirically over a time period of length T hours , so we now introduce T into the definition of B . Denote density by B ( T;D , E ) and re-write i as i ( θ , T ) to account for the change . The time-dependent optimal combination , θopt ( T ) , then satisfies ( 6 ) It follows by definition that θopt ( T ) and θsyn are equal when T = 0 and are therefore also close for small T , Equation 3 describes the rate of divergence between the two . If we define the dimensionless interaction profilethe degree of interaction , I ( T ) , is given by the mean interaction taken over the relevant drug combinations:Negative I ( T ) denotes synergy , positive I ( T ) denotes antagonism . A measure of the convexity and concavity of i ( θ , T ) obtained by fitting a quadratic , , can be used to assess the drug interaction . Significant positivity ( obtained using a t-test ) of α indicates synergy , negativity indicates antagonism; Section 7 in Text S1 gives further information on the use of this test . If the density data is significantly nonlinear as a function of θ , meaning α≠0 , the fitted quadratic can be used to robustly estimate the drug proportion that maximises bacterial density at each time . This proportion is given by one of θ = 0 , 1 or −β/ ( 2α ) depending on which value is the lowest of q ( 0 ) , q ( 1 ) or q ( −β/ ( 2α ) ) . Provided −β/ ( 2α ) lies between 0 and 1 , an approximate upper bound on the confidence interval for this optimal value can be found from a t-test that returns confidence intervals for α , β , and γ . Throughout we will refer to the test described in this paragraph as the ‘α-test’ and it is implemented using the regression facilities in the Statistics Toolbox of MATLAB . | We take an evolutionary approach to a problem from the medical sciences in seeking to understand how our knowledge of rapid bacterial evolution should shape the way we treat pathogens with antibiotic drugs . We pay particular attention to combinations of different drugs that are purposefully used to produce potent therapies . Textbook orthodoxy in medicine and pharmacology states one should hit the pathogen hard with the drug and then prolong the treatment to be certain of clearing it from the host; how effective this approach is remains the subject of discussion . If the textbooks are correct , a combination of two antibiotics that prevents bacterial growth more than if just one drug were used should provide a better treatment strategy . Testing alternatives like these , however , is difficult to do in vivo or in the clinic , so we examined these ideas in laboratory conditions where treatments can be carefully controlled and the optimal combination therapy easily determined by measuring bacterial densities at every moment for each treatment trialled . Studying drug concentrations where antibiotic synergy can be guaranteed , we found that treatment duration was crucial . The most potent combination therapy on day 1 turned out to be the worst of all the therapies we tested by the middle of day 2 , and by day 5 it barely inhibited bacterial growth; by contrast , the drugs did continue to impair growth if administered individually . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"control",
"theory",
"functional",
"genomics",
"applied",
"mathematics",
"population",
"genetics",
"genome",
"sequencing",
"mutation",
"systems",
"ecology",
"mathematics",
"theoretical",
"ecology",
"infectious",
"diseases",
"biology",
"evolutionary",
"theory",
"nonlinear",
"dynamics",
"infectious",
"disease",
"modeling",
"microbial",
"ecology",
"ecology",
"natural",
"selection",
"genomics",
"evolutionary",
"biology"
] | 2013 | When the Most Potent Combination of Antibiotics Selects for the Greatest Bacterial Load: The Smile-Frown Transition |
With increased global attention to neglected diseases , there has been a resurgence of interest in eliminating rabies from developing countries through mass dog vaccination . Tanzania recently embarked on an ambitious programme to repeatedly vaccinate dogs in 28 districts . To understand community perceptions and responses to this programme , we conducted an anthropological study exploring the relationships between dogs , society , geography and project implementation in the districts of Kilombero and Ulanga , Southern Tanzania . Over three months in 2012 , we combined the use of focus groups , semi-structured interviews , a household questionnaire and a population-based survey . Willingness to participate in vaccination was mediated by fear of rabies , high medical treatment costs and the threat of dog culling , as well as broader notions of social responsibility . However , differences between town , rural and ( agro- ) pastoralist populations in livelihood patterns and dog ownership impacted coverage in ways that were not well incorporated into project planning . Coverage in six selected villages was estimated at 25% , well below official estimates . A variety of problems with campaign mobilisation , timing , the location of central points , equipment and staff , and project organisation created barriers to community compliance . Resource-limitations and institutional norms limited the ability for district staff to adapt implementation strategies . In the shadows of resource and institutional limitations in the veterinary sector in Africa , top-down interventions for neglected zoonotic diseases likes rabies need to more explicitly engage with project organisation , capacity and community participation . Greater attention to navigating local realities in planning and implementation is essential to ensuring that rabies , and other neglected diseases , are controlled sustainably .
Rabies has been known since antiquity as one of the most feared human diseases [1]–[3] . Today , it remains a significant albeit neglected disease , causing some 55 , 000 deaths each year , predominately among children and the rural poor in Asia and Africa [4]–[6] . Transmitted by saliva from the bite of an infected animal , the rabies virus invades the central nervous system and , in the absence of post-exposure prophylaxis ( PEP ) , is fatal once clinical signs appear [7] . Symptoms can be nonspecific but often include hydrophobia , hypersalivation , respiratory difficulties , biting and aggression . Although all mammals can be infected , the vast majority of human rabies cases are caused by domestic dogs [8] . Canine rabies has been eliminated from most industrial economies . In Great Britain , this was achieved in 1902 through a combination of dog licensing , muzzling , culling , tracing movements of rabid dogs and their contacts , and strict quarantine , which continues to be upheld by “pet passports” [3] . However , dog vaccination is now regarded as the most effective control strategy combined with secondary roles for population control , movement regulations and the promotion of responsible dog ownership [8]–[10] . There is a strong economic argument for dog vaccination , as eliminating infection from dogs should reduce the demand for costly PEP [11]–[12] . Yet dog vaccination remains under-prioritised by most developing countries with competing health issues and limited resources . Perceptions held by policymakers are that operational constraints ( a lack of knowledge about the dog population , inadequate resources and wildlife transmission ) are barriers to vaccination [8] . These perceived barriers may be “overstated and erroneous” [9] as a number of successful initiatives have been implemented . Since the 1980s , for example , a combination of intensive canine vaccination and surveillance efforts in Latin America has shown dramatic progress [13] . However rabies has been increasing in parts of Asia and Africa and remains widespread in over 80 countries [14] . Recently , a number of initiatives have been undertaken [15]–[18] , bolstered by new elimination targets set by the World Health Organisation [19] . Rabies is endemic in Tanzania with an estimated 1 , 500 deaths each year [20] . Two decades of research in northern Tanzania has generated important epidemiological insights while demonstrating that the disease can be controlled [5] , [11] , [21]–[22] . Tanzania was among three countries selected by the WHO for large-scale rabies elimination demonstration trials between 2009 and 2013 funded by the Bill and Melinda Gates Foundation ( BMGF ) ( see: http://www . who . int/rabies/bmgf_who_project/en/index . html ) . This represented a shift from a localised research project towards an integrated government programme managed by the WHO country office and implemented by government ministries . This ongoing project stretches over 28 districts in Dar es Salaam , Lindi , Morogoro , Mtwara , Pwani and Pemba regions with a diverse population of over 6 million people and an original estimate of 400 , 000 dogs . The project comprised annual free dog vaccination campaigns , free supplies of PEP to rural health clinics , and improved surveillance for five years in each district . After the project , dog vaccination was to be institutionalised within the Tanzanian government , who would then pay for maintaining successes and scaling-up activities to other areas of the country as part of a sustainable country-wide programme . The project aimed to demonstrate the feasibility of rabies elimination in a sub-Saharan African context with a strong focus on country ownership , envisioned to help catalyse the development of national programmes in other countries . To successfully eliminate rabies , vaccination must reach at least 70% of a dog population over consecutive years [14] . Vaccination rates lower than 30% are considered a “waste of resources” [8] . Vaccination coverage declines rapidly in dog populations with high turnover rates [23] . Most dogs in Africa are owned by a family but are free-roaming and generally quite young; some studies show that half of dogs are less than one year of age [24]–[27] . Validated estimates of dog populations are mostly lacking; a recent study in Iringa district , Tanzania showed that the dog population was six times larger than official estimates [25] . However , such estimates are essential for planning successful mass dog vaccinations . Despite the feasibility of rabies elimination , most vaccination efforts in Africa have failed to achieve high levels of coverage [8] . Interventions are clearly influenced by local dog ownership practices . For example , attitudes towards dogs and the ability and willingness of owners to handle their dogs; the location of vaccination points; and the extent of information dissemination and knowledge of rabies have all been shown to influence compliance [15] , [28]–[30] . Dog owners have not been willing to pay the full costs of vaccination , indicating that rabies control should be considered a public good [31] . Central points are not sufficient in some settings; despite higher costs , house-to-house strategies were needed to achieve 70% coverage in more dispersed pastoralist communities in Northern Tanzania [21] . Whilst dog-owner characteristics are important in understanding project outcomes , the capacity and working norms of implementing organizations also play central mediating roles . Although planned at the central level , most campaigns are delivered through ( sub- ) district-level livestock field officers who mobilise dog owners to attend central vaccination points . Due to the legacy of structural adjustment on the veterinary sector , the state's capacity in animal health is generally limited in much of Africa [32] . Large and remote geographical areas together with low salaries , insufficient resources and rigid bureaucratic norms can further inhibit such campaigns which depend , to a large degree , on adapting strategies to fit community needs [33] . Hence there are risks that new large-scale rabies control programmes in Africa will encounter fairly stereotypical challenges of “top-down” public health interventions in developing countries , known to overlook critical social , cultural , political and economic contexts that mediate effectiveness [34]–[36] . The ethnographic literature is replete with examples of how otherwise efficacious biomedical interventions fall afoul due to divergences in , among other things , issues of power , knowledge , interests and social norms between different social groups; for example , polio in Nigeria [37] , schistosomiasis and soil-transmitted helminths in Uganda [38]–[39] and lymphatic filariasis in Tanzania [40] , tuberculosis in Nepal [41] and avian influenza in southeast Asia [42] . Furthermore , recent work recognising the interrelationships between social and ecological contexts and drivers in infectious disease control ( i . e . One Health and EcoHealth ) [43]–[44] as well as the complexities of fostering equitable access to health technologies for the poor ( issues of acceptability , adequacy , affordability , availability and organisational architecture ) [45]–[46] have highlighted the need for programmes to better understand and engage with the key “effectiveness determinants” that mediate outcomes . It is increasingly imperative , therefore , that Neglected Tropical Disease ( NTD ) research explores the perceptions and responses of communities and frontline health and veterinary workers to interventions in order to critically analyse their impact and help tailor programmes for sustainability [47]–[48] . Previous studies examining dog vaccination coverage have been largely quantitative and focused on demographic and spatial factors affecting coverage as well as behavioural characteristics of individual dog owners . To date , there have been no studies detailing how the perceptions , behaviours and contexts of different local actors influence such campaigns , as promoted by actor-oriented perspectives in sociology and anthropology [36] . This article reports the first mixed method anthropological study on canine vaccination in Africa , focused on the predominately rural areas of Kilombero and Ulanga districts in Southern Tanzania .
Research was conducted in Kilombero ( 14 , 918 km2 ) and Ulanga ( 24 , 560 km2 ) districts in Morogoro region , Southern Tanzania , during the dry season from May-August 2012 ( Figure 1 ) . These districts are surrounded by the Udzungwa Mountains National Park and the Selous Game Reserve and are roughly divided by one of the largest wetland areas in Africa , the Kilombero Valley ecosystem . The rainy season begins in early November and ends in May . Occasional dry spells from December to March ameliorate flooding that disrupts road transport in the Kilombero Valley during the rainy season . A large diversity of ethnic groups have come to inhabit the area during several historical migrations , include the Ndamba , Pogoro , Mbunga , Bena , Ngoni , Ngindo and Hehe , who speak their local languages as well as Kiswahili [49] . People depend heavily on the natural environment for water , wood , pasture , bush-meat and farming . The economy of the Kilombero Valley is structured around the farming of rice and maize , livestock keeping , small business , fishing and casual labour . There are also a few large plantations of sugarcane , rice and teak and other formal employment in urban areas , including the district centres Ifakara and Mahenge . Religious affiliation is roughly 40% Muslim and 60% Christian . In 2006 some 657 , 899 people resided throughout 146 villages within the two districts , with a much higher population density in Kilombero than Ulanga [50] . The area lacks tarmac roads outside the district capitals as well as easy access to a national highway ( travel to Ulanga requires the use of a motorised ferry connected to Kilombero ) , which has certainly helped maintain the areas relative economic and political marginalisation despite its abundant natural resources . Importantly , dog vaccinations had been conducted in Kilombero and Ulanga for two years prior to the WHO/BMGF project by local researchers following a rabies outbreak in 2007 . This was unique among the 28 WHO/BMGF project districts , which had only commenced district-wide vaccinations in 2010; hence our two study districts offered an opportunity to learn lessons about how district teams adapted over time to vaccination campaigns . Implicitly , we assumed that this would translate into improved planning , education , engagement with community needs and understanding of the local dog population as compared to other districts in the project . The study involved five phases of fieldwork ( Figure 2 ) . The first involved focus group discussions ( FGDs ) with separate groups of women and men ( between 6 to 15 people ) in 16 villages ( 8 villages in each of the two districts ) . These participants were selected in collaboration with the village office to contrast differences in socio-economic status , deliberatively mixing wealthier , middle and poorer participants and those with and without dogs . Semi-structured interviews ( SSIs ) were also individually conducted with each village leader to clarify details and explore related topics . These interviews and focus groups explored people's knowledge and experience of rabies , attitudes and opinions of the vaccination campaign and dog management practices and attitudes towards dogs . These villages were chosen to incorporate a range of estimated vaccination coverage and known rabies cases ( provided by district officials ) and included those villages with the most known cases of human rabies and those with no reported cases . Villages were also selected to maximise differences in demographic , cultural and geographical variation , as based on the knowledge of local researchers . In the second phase , semi-structured key informant interviews were conducted with senior district officials in the medical ( 3 ) , veterinary ( 3 ) and agricultural sectors ( 2 ) as well as with 11 livestock field officers responsible for vaccination . The third phase involved selecting six of the 16 villages originally visited for more in-depth study . Careful attention was given to maximising common variations that emerged from the focus group data , including differences in coverage , rabies cases , livelihood patterns , social characteristics , geography and dog density and management . A population-based survey was conducted in these villages where enumerators visited every household to gather data on the human and dog population as well as vaccination status of dogs and reasons for non-compliance . A total of 6 , 157 households were found and spot checks of 20 households per village were conducted to verify the accuracy of this data . Fourth , a detailed household questionnaire ( HHQ ) with both open and closed ended questions was done with approximately 20 dog owners in each of these six villages ( n = 113 ) . Most rural villages were large and dispersed with upwards of 10–20 km in diameter and composed of four to eight sub-villages; hence questionnaire administration was divided equally between the different sub-villages ( ranging from four to eight ) of each village where an effort was made to seek out households in the most remote and dispersed settlement areas . This questionnaire explored livelihood characteristics , dog management , disease knowledge and attitudes towards vaccination . Since residents from remote sub-villages were often few in the initial focus groups , clarification of their experiences was necessary and one focus group was then done with community members ( half were male , and half were mixed gender , groups ) in the most remote areas of each of the six villages on similar topics to those described above . Lastly , five key informant interviews were done with researchers involved in rabies control in Tanzania to better contextualise the study . For qualitative data collection verbal informed consent was obtained from each research participant while for quantitative data collection written consent was used . All data collection , except for key informant interviews , was conducted in Swahili and translation from and into English was done . All questionnaire data was entered and analysed using Excel ( Microsoft Office Excel 2007 ) . Qualitative data was entered into Microsoft Word and analysed manually . Ethical clearance was obtained from Sokoine University of Agriculture in Tanzania ( Ref: RPGS/R/8VOL XI ) .
As an intervention , 70% coverage of the dog population is needed over consecutive years for rabies vaccination to be successful , making a good knowledge of the dog population essential to planning and estimating coverage . Interviews with the District Veterinary Officers ( DVOs ) of the two districts showed that the dog population was not well documented . Available data from Kilombero included the 2002 census that reported 21 , 941 dogs and an informal estimate given by the DVO that this had “now gone up to about 29 , 000 dogs . ” For Ulanga , this included a 2009 census that showed 7 , 385 dogs . Based on the 2006 human census estimates , this would give a human-dog ratio of 12 . 3∶1 in Kilombero and 28 . 7∶1 in Ulanga . These are both relatively low estimates compared to other published studies [9] . Other studies in Tanzania in both coastal and inland regions estimated a human-dog ratio of 14∶1 , albeit inland rural areas ( like Ulanga and Kilombero ) had a much higher ratio [30] . Work in the Serengeti among pastoralist and agro-pastoralists showed a ratio of 6 . 3∶1 [11] and 7 . 3∶1 [21] , while a recent study in a Tanzanian city ( Iringa ) found a 14∶1 ratio , six times larger than the official district records [25] . Dog registers kept in the DVOs office indicated the name of the owner of each vaccinated dog , allowing for tentative estimates of coverage . For the DVOs , this contributed to estimates of coverage that were far higher than was likely the case: the DVO of Kilombero cited 75% then reduced it to “at least more than 50% for sure” with some reluctance , while the DVO of Ulanga stated that “at least 90% of the dogs in the district were vaccinated , certainly not less ! ” However , that rabies was still present ( discussed below ) , albeit reduced from the 2007 outbreak levels , should have been indicative of a much lower coverage , at least for Ulanga . This is especially the case given that rabies oscillates between endemic and outbreak scenarios [51] . Using the official dog population estimates provided by the DVOs and the 2011 vaccination data from their offices , vaccination coverage for 2011 was 40 . 5% in Kilombero and 102% for Ulanga , with lower figures for 2009 and 2010 ( see Table 1 ) . Unlike with Kilombero where routine vaccination was also done , dogs were only vaccinated in Ulanga during the campaign as the district lacks the necessary cold chain outside the district capital . In discussions with government officials and villagers it became clear that there were very different assessments of how successful the vaccination campaigns had been . Apart from the low coverage in Ulanga in 2010 ( explained below ) , government perceptions emphasised that coverage had been increasing in parallel with the experience of the extension officers , the addition of more central points , the involvement of teachers , nurses and doctors during the campaign , and greater practice and trust with dog-owners . The high coverage reported by DVOs was reiterated by the 11 interviewed livestock field officers ( LFOs ) , most of who had been involved in all three or four campaigns . Despite some scepticism that 70% of dogs had been vaccinated , not one believed less than 50% had been vaccinated with most placing the estimate at 60% and some more than 80% . In contrast , focus groups and interviews with community members emphasised the small proportion of vaccinated dogs , placing their own unofficial estimates between 25 to 50% coverage . Understanding vaccination coverage requires considering the various links between livelihoods and dogs in Kilombero and Ulanga , which varied greatly between social groups with important implications . While there were other common uses for dogs ( hunting , companionship , symbols of wealth , to ward off spiritual forces and act as capital assets when selling puppies ) by far the most important involved security −97% of questionnaire respondents stated so . However , the particularities of how dogs were used for security , the human-dog relationship and how dogs were managed differed between cattle keepers ( both agro-pastoralists and pastoralists ) , rural farmers and town residents . For farmers ( the majority of the rural population ) , dogs represented a “line of defence” between crops and certain destructive wildlife . For example , 86% ( n = 97/113 ) of questionnaire respondents claimed to suffer from varying degrees of wildlife encroachment on their farms . While elephants and buffalo could cause major damage , these were rare and monkeys and baboons were the major problems where they , in the words of one angry woman , “finish off a large portion of my crop in one day and enjoy harassing our maize the most . ” This was further impacted by the geography of the farm relative to the homestead , forests and wetlands . Following the villagization programme of the post-independence era in Tanzania [52] as well as the need to cultivate rice in the wetlands , homesteads were often far away from farms . During the growing season , farmers either migrated from the village to a small makeshift hut for a few months or commuted daily from their homes , with many taking their dogs with them . Wildcats , mongooses and jackals were known as thieves of chickens and chicken eggs and dogs were also commonly kept to protect them . Reasons for keeping dogs were different for livestock keepers . For Masai and Mang'ati pastoralists and the agro-pastoralist Sukuma ethnic group , dogs were used to guard cows , goats and other livestock during grazing and in the cowshed at night from thieves as well as wild dogs , jackals , hyenas and the occasional lion and leopard . In small groups sometimes with hundreds of cattle , young men migrated between the village outskirts and the wetlands and forests following pasture and waterholes during the dry season . These were generally either a short one or two hour walk away from the home ( if routes did not infringe on farmland ) or large distances of upwards of 20 km or more . Women , children and elders would remain resident in the village during these migrations , most often in remote and dispersed sub-villages far from main access routes . Pastoralists were considered ( and observed ) to own many more dogs than farmers and their dogs were also bigger , more aggressive and more loyal and alert . Long migration routes as well as cultural determinants ( i . e . emphasising a “warrior” attitude , common to these pastoralist groups ) cultivated closer bonds between dogs and male ( agro ) -pastoralists than with most sedentary farmers . This contrasted with the small or large town centres dispersed throughout the area where thieves were the main rationale for dog ownership . There were a number of reasons given for why dog populations were considered far smaller in more densely populated areas: land owners not permitting the keeping of dogs; the higher chance that town dogs would cause conflict by biting people in the street; an idea that urban residents were more “educated” and would keep fewer dogs; and urban residents reporting that they practiced “proper” Islam that restricted the keeping or touching of dogs . According to certain Koranic rules , these groups emphasised that physical contact with a dog ( saliva and fur ) would make someone spiritually unclean ( especially before prayers ) . For these reasons , Muslims in towns stressed that , although they could keep dogs , they had to “treat them well as Mohammed said…and have them only for a specific purpose . ” Regardless of religion , urban dogs were believed to be better cared for and more likely to be vaccinated than dogs in rural areas , with a few confined to their household ( unlike the vast majority of dogs that were free roaming ) . Therefore , differences in livelihood patterns ( and their culturally-embedded dynamics ) between town , farmland and pastoralist systems influenced the human-dog relationship and the spatial distribution of dogs in Kilombero and Ulanga . Utilitarian value tended to mediate and dictate dog management rather than purely culturally-defined beliefs and practices . This clearly impacted vaccination coverage rates: villages that believed vaccination coverage was highest were from more urban areas situated along main roads but with fewer dogs , whilst lower coverage estimates were given in those villages in more remote areas , known to have higher dog populations . Local knowledge of rabies also revealed a general perception of low vaccination coverage , reflected in understandings of rabies epidemiology , experiences of rabies cases and attempts by village leaders to institutionalise “village laws” in order to address non-compliance . Rabies was linked to its Kiswahili name Kichaa cha Mbwa ( madness of dogs ) and widely known as a fatal disease of dogs and humans that affected the brain , was transmitted by animal bites and prevented by dog vaccination , similar to a recent large-scale questionnaire study in Tanzania [53] . Aside from this basic knowledge , rabies was considered an “outbreak disease” , understood in relation to four interrelated beliefs . First , it was a disease of “dirty dogs” caused by neglected ( but owned ) free roaming dogs that spread the disease due to poor animal welfare and poverty . This narrative emphasised that although most farmers and town residents claimed to own dogs for security , this was often an assumed rather than actual use . Many dogs were considered lazy , not aggressive enough , unable to be trained and always away from home looking for food or a dog of the opposite sex . They lacked a clear utilitarian value , which in turn fostered “negligent owners” who did not care for their animals and , therefore , facilitated the spread of rabies . In the words of one village leader , “living as we are in this farming environment [as poor farmers] , dog owners keep dogs without a purpose and do not care about them so they move all over the place…and this is how they catch rabies . ” The second common narrative involved the idea that rabies had never been a problem in the Kilombero Valley until the migration of Masai and Sukuma from northern Tanzania imported rabies as they moved into the area in the late 1990s , which strengthened animosity between farmers and ( agro- ) pastoralists in certain areas [54] . Third , rabies was believed to spread from wildlife to dogs , facilitated by farmers , hunters and pastoralists living near game reserves and national parks and influenced by seasonal changes in rainfall affecting the movement of carnivores . Lastly , rabies incidence was considered to increase during the harvest period in June and July corresponding with the mating season . The majority of people approved and understood the role of canine vaccination . Differences between biomedical and local understandings , known to lead to community resistance to other human and animal vaccination programmes [37] , [55] , were largely absent . Although rumours that the vaccines were killing dogs and that the campaign was a government dog culling programme had been widely disseminated during the 2008 and 2009 campaigns ( before the WHO project ) , these concerns had abated with time and side effects to vaccines ( real or perceived ) were rarely mentioned . Part of this had to do with the high level of awareness about rabies , underpinned by local experiences of human cases . Although open to error , focus group participants and village leaders identified ( with detailed symptoms and related circumstances ) a total of 59 suspected rabies death cases in the 16 study villages within memory , most ( 45 ) reportedly from 1995 to 2008 , but with four deaths identified in 2012 ( the year of field research ) . While most were from dogs , there were a few attacks from jackals and wild dogs . This would give an average of 3 . 2 cases per year ( 1995–2008 ) in these 16 villages ( population 30 , 143 ) , implying 10 . 7 cases/100 , 000 people; much higher than the 4 . 9/100 , 000 estimated for the country as a whole based on active surveillance in Northern Tanzania ( this difference can be attributed to the fact that our selected villages included those most affected by the outbreak between 2007–2008 ) [20] . Contact tracing as part of a related research project ( where researchers follow up all reported suspected rabies bite cases ) showed 30 deaths in the two districts since 2007 , with most prior to 2009; in turn , hospital records between 2009 to mid-2012 showed 478 bite victims of suspected rabid animals divided equally between the two districts , with only 2 reported deaths ( Unpublished data ) . These local accounts of having neighbours and relatives die from rabies or have to seek treatment after being bitten by a dog generated a significant degree of fear and apprehension . This clearly motivated many households to comply with vaccination . Asked if they would prefer acquiring HIV/AIDS or rabies , 33% of questionnaire respondents picked HIV/AIDS while 14% could not choose between the two . While people mentioned hydrophobia , muscle spasms and nervous twitches , they stressed that respiratory symptoms made victims “bark” like the animal that had transmitted the disease: rabies made people “act like wild animals” and “die like mad dogs . ” They became “demon-possessed” , started to “bite everything” and become “so strong like the animal that bit them . ” Furthermore , access barriers to treatment ( high costs and inadequate access to medicines and health services more generally ) drove community fears . As one woman stated , “For rabies , if you are bitten today and cannot get treatment , which is so common here , tomorrow you die like an animal” ( Focus group participant , Sanje village , Kilombero ) . This level of fear drove communities to attempt to institutionalise two different “village laws” in order to increase compliance with vaccination and deal with suspected rabid dogs and bite victims . In response to the 2007 outbreak and recent vaccination campaigns , most villages had established local bylaws indicating that dog bite victims should be financially compensated for medical costs by the dog owner if the dog was not vaccinated; albeit compensation was never guaranteed . Some never pressed their neighbours for payment , others were not able to identify the dog owner , and others were not able to prove ( in the village court ) that the accused dog actually belonged to the owner ( given the lack of records ) or was not vaccinated ( certificates could be used interchangeably between dogs ) . Second , there were various endogenous attempts to standardise dog culling after vaccination , considered an ethical and effective method of rabies control at the village-level ( but in no way promoted by the WHO project ) . In many villages killing unvaccinated dogs was considered a “district law” with support from livestock field officers; albeit the passing of the Animal Welfare Act ( 2008 ) made this law ambiguous . The most common suggestion to improve coverage was for the village office to require dog owners to register their dogs so that after a vaccination campaign , a grassroots “local committee” could move house-to-house eliminating unvaccinated dogs ( evident by the lack of a new collar and the vaccination certificate ) . This was often done by villagers themselves in haphazard ways that led to protests from dog-owners . Responses to dog bites ( despite many caused by aggressive dogs , bitches with puppies , dogs defending their homestead from strangers or provocation ) were always treated as suspected rabid cases and involved quickly killing the dog , and often provoked a spontaneous dog culling spree . The importance of strengthening these two endogenous attempts to enforce dog vaccination was ubiquitously emphasised , reflecting local perceptions that the rabies control project was achieving low-levels of coverage . During focus groups and interviews , the relationship of rabies to “negligent” dog owners , pastoralists , wildlife and seasonal variation quickly veered into discussions about how vaccination campaigns was not sufficiently addressing what were considered key points for controlling the virus; there was a need to better prioritise targeting households bordering wildlife populations , synchronise vaccination with the farming season and pastoralist migrations , and motivate the many “negligent dog owners” through recourse to village laws and punishments , supported by district authorities more systematically . But how many dogs were truly being vaccinated ? Given the divergent views of government officials and villagers , there was a need to generate more robust estimates of the dog population and vaccination coverage; hence , we carried out a population-based survey in six selected villages . The survey showed that out of a total of 6 , 157 households and 30 , 143 people , there were 1 , 311 dog-owning households ( 21% of households ) and 3 , 056 dogs ( Table 2 ) . This included 2 , 414 dogs older than one year and 642 dogs less than one year . While this gave a total human-to-dog ratio of 9 . 86∶1 , this was highly skewed following local knowledge that the dog population was predominately in rural and remote areas . The more urban villages ( or towns ) of Mwaya and Chikwera had a human-to-dog ratio of 31 . 4∶1 and 64∶1 while the rural villages of Mofu and Namhanga had ratios of 6 . 9:1 and 5 . 8∶1 . However the low population in Mwaya was also a consequence of mass dog culling campaigns that had taken place in 2008 and 2010 in response to human rabies cases . This variation was equally pronounced within each of these villages . For example , sub-villages bordering forests in Machipi and Mwaya had a much higher human-to-dog ratio than other areas . Likewise , the sub-villages with pastoralists in Namhanga and Signali had double , and in Mofu village more than 10 times , more dogs compared to other sub-villages but with relatively equivalent human populations . This showed that the dog population was highly skewed even within individual villages , based on surrounding ecological characteristics that influenced dog utility . Furthermore , the population-based survey also confirmed the low coverage emphasised by community members . In total , only 769 dogs ( 25% of the canine population ) had been vaccinated in 2011 , whereas 2 , 287 dogs ( belonging to 923/1 , 311 households ) had not been vaccinated . If the 642 dogs born since the vaccination campaign ( 21% of the dog population ) are excluded , coverage rises to 32% of the mature dog population . The immunised population is slightly lower given the small percentage of stray dogs; however , this is a relatively negligible population given scarce food resources , estimated at 3–5% in rural northern Tanzania and 1% in urban areas of Iringa , Tanzania [11] , [21] , [25] . As with dog density , vaccination coverage also varied between villages ( Table 2 ) with the highest coverage in both Machipi and Chikwera villages and lower coverage in Mofu and Mwaya . Importantly , dogs in the low coverage villages of Mofu and Namhanga together accounted for 71% of the total dog population of the six villages ( with 2 , 175/3 , 056 dogs ) due to settlements of pastoralists and remote farmers in a number of sub-villages , which were far from main access roads . In contrast , the two villages with highest coverage rates included a large town with only 65 dogs ( Chikwera ) and a village ( Machipi ) relatively close to the district capital in Kilombero . While most people understood the role of canine vaccination , interrelated geographic , social and operational factors created a number of important access barriers . In the population-based survey , reasons given by the 750 dog-owning households ( with dogs born before the 2011 campaign and considered eligible for vaccination ) for non-compliance included ( in descending order of importance ) : not being aware that the campaign was taking place ( 23% ) , having a central point too far from their homestead ( 16% ) , not being able to find their dog ( 14% ) , not being available that day ( 12% ) , the vaccine having run out ( 10% ) , having the dog run away during transport or at the central point ( 10% ) , not being aware of the importance of vaccination ( 7% ) , not being able to catch the dog ( 6% ) , having a young puppy or pregnant female ( 2% ) , a perception that the vaccine has side effects ( 1% ) and having just recently moved to the area ( 0 . 2% ) ( Figure 3 ) . However , understanding how and why these various barriers existed requires triangulating this with qualitative data .
The feasibility and cost-effectiveness of rabies control and/or elimination through canine vaccination has been well documented , with some noted successes from developing country contexts [9] , [11] , [13] , [18] . However there are clearly challenges in mobilising resources for canine vaccination as well as operational barriers that inhibit success in many contexts . With renewed global attention to rabies following advocacy efforts by the NTD community , there is a need to think critically about how local realities intersect with technical solutions; how should we think about the challenges of dog vaccination for rabies and , importantly , how can large-scale canine vaccination projects navigate local social and ecological complexities in resource-limited settings ? Much recent work in the field of sustainable development and global health ( including that of many anthropologists ) has emphasised the importance of understanding interventions from the perspective of community-equity effectiveness and using transdisciplinary approaches rather than narrowly emphasising the efficacy of scientific tools and strategies [33]–[48] , [57]–[58] . Effectiveness has been conceptualised as a “step ladder” where different variables ( at multiple levels ) have lesser or more impact on outcomes depending on social , cultural , biological , economic , political and ecological contextual factors [59] . Analytically investigating these “effectiveness determinants” is deemed essential to understand their multiplicative effects . Intervention planners , therefore , are encouraged to identify and engage with high-level determinants , enabling factors and local capacities ( that act as essential nodes ) in order to move away from managing risk to building resilience and understanding interventions as “complex systems” [44] , [60] . Exploring the implementation and community response to a WHO-coordinated canine rabies elimination project in two southern districts of Tanzania , this article has presented ( to our knowledge ) the first anthropological study of a contemporary dog vaccination programme in a resource-poor country . In the absence of credible estimates , a population-based survey in six selected villages showed that 25% of the dog population had been vaccinated in 2011 . The survey quantified what was general knowledge among the village population – that the campaign had achieved coverage well below the 70% target due to a number of interrelated social processes , geographical characteristics and challenges in project implementation . Furthermore , while it is difficult to extrapolate the findings of this study to the wider WHO project area , many key informants believed that Kilombero and Ulanga , due to its prior experience with mass dog vaccinations , achieved relatively high levels of vaccination coverage , suggesting that the difficulties encountered here were not unique . But what were the most important bottlenecks to the canine vaccination project in these two rural districts that had the greatest leverage on mediating intervention effectiveness , and therefore should be most emphasised and reflected on for future vaccination campaigns in Tanzania and elsewhere ? At the community level , there were clear spatial differences in dog distribution driven by the variable dog keeping practices of rural farmers , town residents and ( agro- ) pastoralists . While dogs played important roles that were embedded within local livelihoods , there were differences between conceived uses and actual ones . Many dogs used “for security” were actually poorly fed and maltreated with little or no clear role in the household . Awareness of rabies , at least on a basic level , motivated people to participate in rabies control out of fear of “dying like a mad dog” as well as , to varying degrees , having their dog culled and being held responsible to pay for someone's medical treatment . Equally important were broader notions of social responsibility that reflected much broader divisions within these communities about the willingness to control diseases that were perceived to be relatively rare . Some people in these predominately rural geographies themselves under-prioritised ( or neglected ) the importance of rabies control given the multitude of other challenges in their daily lives . The widespread emphasis on the need for local bylaws to punish dog owners who did not vaccinate their dogs and monitoring of vaccination status by the village office was a general expression of a desire to motivate ( and coerce ) non-compliant “negligent” dog owners . Given the difficulties of behavioural change in resource-limited settings [61] , there is surely an important role to sustain education campaigns to help increase and facilitate prioritisation at the village-level over the long-term , with a possible role for dog registration . However , barriers to vaccination did not rest solely , or predominately for that matter , with communities . The rabies elimination project suffered from stereotypical challenges of “top-down” public health programmes . There were critical gaps in communication between central government authorities , district officials , field staff and the target population that were structured by existing bureaucratic procedures , social norms and an over-emphasis on technical solutions . In both districts , an underestimation of the dog population increased what was found to be an erroneous perception of success . The dog population was not geographically uniform but heavily skewed , found largely in more remote areas bordering forests and the outskirts of pastoralist villages , than the more accessible towns or areas with easy access routes . These relationships found expression in local understandings of rabies epidemiology – related to pastoralist migration and wildlife interaction – which were not well incorporated into project planning . These operational challenges were exacerbated by the long-term effects of structural adjustment policies in the veterinary sector in Tanzania that have significantly reduced the capacity of the state to deal with animal health [32] . This found expression in the negative attitudes of most villagers towards their local livestock field officers; the lack of sufficient fuel , vaccines , staff and “promised” salaries; and the perceived inability of district officials to adjust budgets to address local challenges , such as the large geographical area and the need to adapt the timing of vaccination campaigns to fit seasonal specificities ( rainfall and migration ) . A mixture of lack of funds , planning and capacity as well as the government's financial distribution system prevented flexible , context-specific strategies . As a result , the effectiveness of mobilisation , the location of vaccination points and the timing of the intervention were not optimal . Efforts to increase involvement of community members in mobilisation or to adapt vaccination points based on local recommendations were generally limited by capacity and funds . It was not that local district officials were necessarily oblivious to these challenges; rather they felt unable to communicate effectively with those in Dar es Salaam ( Tanzania's capital ) with sufficient power to enable flexibility . Communication channels were top-down and learning from past shortcomings , or putting this learning into practice , was generally limited . Some of these challenges contrast with rabies research programmes ( i . e . work in the Serengeti ) where more capacity and flexibility were believed to have allowed for better targeted campaigns and more community involvement . Between these different geographical , community and organisational dimensions to the vaccination project , this study shows that , despite many endogenous challenges at the level of the dog-owner , issues of capacity , finances and managerial shortcomings severely lowered coverage by preventing field strategies to be adapted to local realities . The major bottlenecks were not with “community compliance” per say but with how intervention strategies navigated the various structural and behavioural factors that mediated access . This shows the need for a more trans-disciplinary and participatory approach in planning , implementing , managing and monitoring and evaluating rabies control programmes . The findings presented here do not suggest that rabies elimination in Tanzania is unachievable; rather , it points to the need to investigate , consider and take seriously local variations and challenges within the project planning cycle . Robust quantitative data on dog populations and vaccination coverage as well as qualitative implementation research are essential for ensuring that project coordinators have a sound understanding of challenges on the ground . These issues , however , are not unique to rabies but rather part of a much larger debate about the nature of vertical health programmes in developing countries , top-down strategies and the relationship between expert knowledge , donor-led development projects and poor populations [33]–[43] . Policy narratives and donor-funded projects are often shaped by presenting “quick-fix” technical strategies that can be easily “scaled-up” from local successes within short time periods [62] . Donors demand results that showcase quick-wins , large impacts and “value-for-money . ” However , there is a tendency to sideline or overlook the scale of capacity building needed as well as the larger bureaucratic challenges involved in fostering “country ownership” and institutionalising equitable and effective interventions within government ministries . Without sound project management that creates feedback loops and adaptive mechanisms between different actors ( paying attention to embedded infrastructure , capacity and community participation issues ) , public health interventions like canine rabies vaccination will have difficulties in navigating local access barriers . Addressing this requires time , leadership , resources , vision and institutional learning to effectively address the legacy of structural adjustment on the health and veterinary systems in developing countries and strengthen the relationship between the central government , district officials , extension workers and communities . Critical gaps between project planners , implementers and communities have also been noted , for example , in other recent studies on Neglected Tropical Disease control in Tanzania [40] , [63]–[64] . Greater realisation that these issues need to be more proactively addressed is shown in contemporary emphasis on implementation research [47] , systems-based approaches to infectious disease control ( i . e . EcoHealth and One Health ) [43]–[44] and the involvement of social scientists in NTD control [48] , [65] . Understanding the context of success and failure , therefore , should be more encouraged by the NTD community if we are to learn from past experiences , propose future strategies and ultimately create more resilient and sustainable programmes , and more healthy communities . An interesting example of how things can change on the ground and the need for flexibility and foresight in implementing a successful rabies elimination programme involves recent changes in dog populations in Kilombero and Ulanga since the end of field research . With the threat of environmental degradation in the fragile Kilombero Valley ecosystem , the government ( with police support ) forcibly evicted over 380 , 000 cattle in late 2012 , likely the majority of pastoralists . As these cattle keepers now migrate to new districts , vaccination coverage in Kilombero and Ulanga will likely increase dramatically , but planning for future campaigns in the wider WHO elimination area will require consideration about where these livestock keepers , and their many dogs , have gone . | Mass vaccination of dogs is the most effective strategy to eliminate dog-mediated human rabies from developing countries . In 2009 , a large-scale elimination demonstration project was funded and coordinated by the Bill and Melinda Gates Foundation ( BMGF ) and the World Health Organisation ( WHO ) in three southern countries , including the United Republic of Tanzania . This paper explores community perceptions and responses to this programme in the districts of Kilombero and Ulanga , Southern Tanzania . The study was based on focus groups and interviews in 16 villages as well as a household questionnaire ( n = 113 ) , a population-based survey ( n = 6 , 157 households ) and key informant interviews ( n = 24 ) . The study showed that fear of rabies , the threat of dog culling and broader ideas of community responsibility drove compliance . However differences in local livelihoods shaped dog ownership patterns and the distribution of dogs in ways that were not explicitly addressed by project strategies . A survey in six villages found that only 25% of dogs had been vaccinated in 2011 . We discuss the operational constraints and problems that lowered coverage as viewed by different actors at the district and village-level . A more explicit engagement with project organisation , capacity and community involvement are needed to address this low coverage . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"science",
"policy",
"biology",
"and",
"life",
"sciences",
"social",
"sciences",
"veterinary",
"science"
] | 2014 | Eliminating Rabies in Tanzania? Local Understandings and Responses to Mass Dog Vaccination in Kilombero and Ulanga Districts |
The prophage is one of the most important components of variable regions in bacterial genomes . Some prophages carry additional genes that may enhance the toxicity and survival ability of their host bacteria . This phenomenon is predominant in Staphylococcus aureus , a very common human pathogen . Bioinformatics analysis of several staphylococcal prophages revealed a highly conserved 40-bp untranslated region upstream of the int gene . A small transcript encoding phage integrase was identified to be initiated from the region , demonstrating that the untranslated region contained a promoter for int . No typical recognition sequence for either σA or σB was identified in the 40-bp region . Experiments both in vitro and in vivo demonstrated that σH recognized the promoter and directed transcription . Genetic deletion of sigH altered the int expression , and subsequently , the excision proportion of prophage DNAs . Phage assays further showed that sigH affected the ability of spontaneous lysis and lysogenization in S . aureus , suggesting that sigH plays a role in stabilizing the lysogenic state . These findings revealed a novel mechanism of prophage integration specifically regulated by a host-source alternative sigma factor . This mechanism suggests a co-evolution strategy of staphylococcal prophages and their host bacteria .
Prophages are viral cellular parasites that integrate into bacterial genomes and co-replicate with host chromosomes . A subset of bacteriophage genomes encodes additional virulence factors , and the production of these virulence factors can enhance bacterial toxicity as well as survival ability in various environments [1] , [2] . Prophages are not rare in the chromosome of Staphylococcus aureus , a widely spread human pathogen . In fact , most clinical isolates harbor at least one prophage [3] . Many staphylococcal phages contain genes that encode virulence factors such as staphylokinase , enterotoxin A , chemotaxis inhibitory protein , staphylococcal complement inhibitor and leukocidin , which greatly enhance the bacterial invasiveness and help to evade host immunity in organic infection [4] , [5] , [6] . The transfer of toxic genes by a lysogenic bacteriophage , or phage conversion , is an important mechanism in the evolution of virulent S . aureus strains [7] . Furthermore , phages participate in the mediation of horizontal transfer of pathogenicity islands and raise intra-strain and inter-strain exchange frequency of toxic genes [8] , [9] . As a member of double-strand DNA viruses , a temperate phage needs to recruit bacterial RNA polymerase with essential sigma factors to initiate its cascade . In S . aureus , only four sigma factors have been identified to date: σA , the housekeeping sigma factor , which directs the transcription of the bulk cellular RNA , and three alternative sigma factors σB [10] , σH [11] , and the newly defined σS [12] . The S . aureus σB protein is closely related to the σB protein of Bacillus subtilis , and is mainly involved in stress response [13] , [14] . The S . aureus σH protein is a homolog of B . subtilis σH , which regulates sporulation-related genes [15] . At the turn of the 21st century , S . aureus genome nucleotide sequences were being completed at a rapid and increasing rate [16] . At least 14 S . aureus strains have been whole-genome sequenced and more are in progress . In addition , many staphylococcal phages have been identified independently and then sequenced as well [17] , [18] , [19] , [20] , [21] , [22] . These sequenced genomes allowed us to comparatively analyze the staphylococcal prophage , one of the most important components of genomic variable regions and which provides numerous virulence factors to host bacteria . The majority of known staphylococcal bacteriophages belong to the order Caudovirales and the size of the phage genomes mainly ranges from 35 to 50 kb . Architectural analysis of these prophage genomes demonstrated identical gene arrangements ( Figure S1 ) . Some prophages share similar open reading frames ( ORFs ) , especially in genes that encode products for the lytic cycle . The integrase genes of these bacteriophages are highly conserved [23] . Some staphylococcal prophages even harbor the same integrase and insert into the same locus on the S . aureus genome . Besides integrase , the excisionase represents divergence among species . Indeed , only some of the staphylococcal prophages contain the xis gene , which encodes excisionase . The rest have a genetic structure called ORF-C , in the opposite direction of int [24] . Here we report a heretofore unrecognized manner whereby an alternative sigma factor is recruited by a staphylococcal temperate phage for the regulation of int transcription . The recognition site for σH is upstream of the ORF of int , which encodes phage integrase . Deletion of sigH resulted in a decrease of phage int mRNA level and an increase of the excised form of prophage genomes under normal growth conditions . The abilities of spontaneous lysis and lysogenization in S . aureus were also affected . These results indicate that S . aureus σH modulates the transcription of phage integrase , stabilizes the lysogeny in the host cell and may further influence the prophage life cycle and correlative bacterial virulence .
We first analyzed 43 staphylococcal prophage sequences available from NCBI GenBank by multiple sequence alignment in segments . From the results of the in silico analysis we found that a small fragment in the 5′ untranslated region ( UTR ) of int was extremely conserved among nearly all the prophages that were compared ( Figure 1 ) , with the only exception of Φ3A , which is int defective [3] . The region was previous mentioned as junction A when compared six S . aureus phages [25] . The conserved region was a 40-bp fragment followed by the typical Shine-Dalgarno ( SD ) sequence AGGAGG and was closely related to int . More interestingly , a previously reported stem-loop structure of 9-base inverted repeat with a 3-base loop ( ATTTAGTACtagGTACTAAAT ) found in several staphylococcal prophages [24] was also adjacent to that region . These data indicated that the region was likely to be a transcriptional regulator binding domain . The DNA sequences upstream of int gene from several other firmicutes harboring prophages and two S . aureus pathogenic islands were also compared ( Figure 2A ) . Similar sequences were identified in Staphylococcus epidermidis prophage ФCNPH82 , ФPH15 , and Staphylococcus haemolyticus ФSH1 , but not in other prophages . It seemed that the conserved region only existed in the staphylococci harboring prophages . Since the results from in silico analysis strongly suggested that the newly defined region was a transcriptional regulation domain for S . aureus prophage int gene , it would be interesting to reveal the transcriptional organization of the integrase . Northern blot assays were performed to determine the transcriptional products for integrases in Ф11 , Ф12 , and Ф13 , respectively . Under the normal lysogenic conditions , two mRNAs encoding phage integrase were produced . The larger mRNAs ( ∼3-kb to ∼4-kb ) were initiated from promoters in the phage immunity region according to their lengths , and the smaller transcripts ( ∼1-kb ) were presumably started from our newly defined region ( Figures 3B and S2 ) . To identify the exact transcriptional initiation site , a primer extension assay was carried out . The signal showed that the start site was just between the conserved region and the translational start codon ( Figure S3 ) . The results confirmed our hypothesis that the conserved region harbored a promoter specific for int expression . However , no typical sequences could be predicted as recognition motifs for neither σA nor stress response sigma factor σB in the newly defined region . Therefore , another protein needs to be recruited to function as a sigma factor . The protein could be either a phage gene product or another bacterial sigma factor . Therefore , we sought to identify this sigma factor in host bacteria . The alternative sigma factor σH was conserved in firmicutes and is known for its sporulation regulation function . Interestingly , staphylococci also express sigH but do not form spores . SigH proteins from staphylococci and other firmicutes were divided into two subgroups according to a previous report [11] . The phylogenetic analysis on sigH orthologs from several firmicutes showed that staphylococcal sigH proteins were separated from the others by a deep node ( Figure 2B ) . Following this , the sigH gene in S . aureus MW2 [26] genome was knocked out to generate ΔsigH ( SUN0802 ) and a decrease in int mRNA levels of both ΦSa2mw and ΦSa3mw was observed by real-time quantitative reverse transcription polymerase chain reaction ( Q-RT-PCR ) . To determine if this phenomenon was strain-specific , another sigH deletion strain ( SUN0806 ) was built up in S . aureus NCTC8325 [27] . Int mRNA levels of Φ11 , Φ12 , and Φ13 were all reduced compared with the wild type ( WT ) ( Figure 4 ) . The sigH gene in RN4220 was also knocked out to generate strain SUN0914 for later experiments . To obtain the complementary strains ( ΔsigHc ) , plasmid pMADsigH was introduced into the above sigH-deficient strains and integrated into the host genomes; the backbone of the plasmid pMAD was then eliminated under the screening at 42°C . Endogenous sigH mRNA could be detected by RT-PCR in both WT and ΔsigHc but not in ΔsigH ( Figure S4 ) . As expected , int mRNA levels in ΔsigHc were recovered compared with ΔsigH ( Figure 4 ) . In the Northern blot assay , the absence of an mRNA signal of about 1-kb for Φ11 integrase was distinguished in ΔsigH compared with the WT and ΔsigHc ( Figure 3B ) . Similarly , the hybridizing signals near 1-kb for Φ12 and Φ13 integrases were not detected in ΔsigH ( Figure S2 ) . The transcriptions of the ORF-C/xis gene on the opposite direction seemed to be unaffected without σH according to our Northern blots ( Figures 3C and S2 ) . However , when the gene locus of int was deleted , transcripts for ORF-C/xis gene could hardly be detected ( Figure 3C ) , and the transcriptional level of cI also decreased ( Figure 3B ) . These results demonstrated that the S . aureus alternative sigma factor σH plays a role in up-regulating the mRNA levels of prophage integrases . It was also interesting to note that the expression of sigH was growth-phase related ( Figure S5 ) . To verify the direct interaction between σH and int promoter region , we conducted σH-directed in vitro transcription by using an amplified DNA fragment from Φ11 , which contained the conserved 40-bp region and part of the int ORF as a template . The core RNA polymerase was pre-incubated with σH and then incubated with the linear DNA template . Transcription was initiated by the addition of an NTP mixture containing [α-32P]UTP at 37°C . The σA-dependent promoter of cro ( pcro ) , an important promoter for transcription in the early period of the phage lytic cycle [28] , and the S . aureus σA protein were chosen as controls . Core RNA polymerase pre-incubated with σH generated the expected signal at the presence of pint , while no corresponding signals were produced by core RNA polymerase , σA-holoenzyme , or sigma factors alone ( Figure 5A ) . By comparison of the intensity of the signals it was also found that the transcription from this σH-dependent promoter was efficient in vitro . To investigate whether σH-holoenzyme could recognize pint and produce transcripts in vivo , a conditional replication , integration , and modular vector pAH125 [29] was used to detect the recognition of the pint with the presence of σH in an Escherichia coli model . A small DNA fragment containing pint was inserted into pAH125 to build pint-lacZ fusion . The constructed plasmid pAH125pint was then transferred into E . coli strain ZK126 [30] and integrated into the bacterial genome at the attP site of the lambda phage with the help of pINTts [31] to obtain TW0901 . Strain TW0902 was later obtained by transferring the plasmid pET22btac-sigH that expresses S . aureus σH protein into TW0901 via electroporation . High activity of β-galactosidase was only exhibited in TW0902 after isopropyl β-D-1-thiogalactopyranoside ( IPTG ) induction ( Figure 5B ) , confirming that S . aureus σH was specific for the int promoter recognition and gene transcription in vivo . Two forms of prophages naturally exist in many lysogenic bacteria , mostly in the integrated form and rarely in the excised form [32] , [33] , [34] , [35] , [36] . In S . aureus , we also observed the coexistence of both integrated and excised prophage DNAs by using a set of specially designed PCR primer pairs ( Figure 6 ) . Primers check-F and check-R were used to amplify the attB sites in S . aureus genome only if the prophages were excised . Primers check-R and check-IN were used to detect the presence of integrated phage genomes . In addition , by using a real-time Q-PCR method with the above primer pairs the excision frequencies of the staphylococcal prophages were estimated . The int gene is required for both integration and excision of temperate phages [37] and some pathogenicity islands [38] as previously reported . We have verified the dual functions of int in the S . aureus prophages by constructing int deletion and overexpression strains . The Φ11 int mutant in S . aureus NCTC8325 ( SUN0818 ) was generated to examine the excision of the prophage . No PCR products were observed using SUN0818 genome as templates with primer pair check11-F and check11-R ( Figure 7 ) , and the excision event could be recovered by introducing a complementary plasmid into the mutant strain ( Figure S6 ) . The Φ11 int overexpression strain was obtained by introducing PLI50Ф11int into RN4220Ф11 . The Φ11 excision frequency in the overexpression strain was decreased and a lower titer was detected in the supernatant of the culture as expected ( Figure S7 ) , indicating that a higher int mRNA level had a role in stabilizing the lysogeny in the host genome . Since sigH directly modulates the expression of phage integrase , we suspected that the defection of sigH gene would lead to a change in the number of excised circular phage DNAs . To test this , the proportion of excised prophage genomes in sigH mutant strains was compared to the WT . The proportions of the excised forms of the genomes of all five prophages tested , including ΦSa2mw , ΦSa3mw in S . aureus MW2 and Φ11 , Φ12 and Φ13 in S . aureus NCTC8325 , were significantly increased . In addition , the prophages excision frequencies were similar in WT and the complementation ( Figure 8 ) . Integration and excision events are usually associated with the life cycle of the temperate phages [39] . We estimated the spontaneous lysis rate by determining the titer of free viral particles in S . aureus NCTC8325 culture at the exponential phase . The supernatants of the respective cultures were incubated with the indicator for adsorption and then plated onto agar plates to form plaques . A higher titer was detected in sigH mutant , suggesting that it was easier to naturally induce the prophages to a lytic cycle in the absence of sigH ( Figure 9A ) . Besides , if the cultures were pretreated with ultraviolet ( UV ) radiation for phage induction , no differences of the titers could be observed between the supernatant of WT , sigH mutant and the complement cultures ( Figure 9B ) . These results implied that sigH did not participate in the phage lytic cycle control . We further investigated the effect of sigH deletion on the lysogenization ability of S . aureus . The susceptible strains RN4220 , RN4220ΔsigH ( SUN0914 ) , and RN4220ΔSigHc were adsorbed by Ф11 particles with the same multiplicity and plated onto agar plates . The survival clones were checked for lysogeny by cross-streak assays . Fewer lysogens were formed in SUN0914 , suggesting that the presence of sigH promoted the lysogenization in S . aureus ( Table 1 ) . These results demonstrated that sigH has an auxiliary role in maintaining the lysogeny in S . aureus by up-regulating the int mRNA level .
Prophages , or lysogenic temperate viruses , provide a large group of virulence factors in bacterial pathogens . In addition , they facilitate the spread of phage genome-encoded or host genome-encoded virulence factors in different bacterial strains or even species [8] . Although the importance of prophage-related virulence in pathogenicity has gradually been realized [40] , much is still unknown about how these virulence factors are modulated . The genomes of lysogenic phages do not permanently stay in its host genome . Occasionally , prophage genomes can be excised from bacterial chromosomes and integrated back later [37] . The mobility of a prophage genome usually reflects the activity of the phage although the detailed mechanism remains obscure . These reversible processes are commonly driven by the int-xis system in temperate phages , and some cofactors may participate as well [41] . The int gene is required for the process of integration , while both int and xis are required for the process of excision [37] . Structural analysis of the integrase family showed that their catalytic boxes are conserved . The major differences in the structures among the members of the integrase family mainly affect the specificity and efficiency of the reactions [42] . Although integrase has dual functions of both integration and excision , a higher int mRNA level usually promotes the reaction of integration that is required for lysogeny . The efficiency of the excision process mainly depends on the protein level of excisionase [39] . A similar regulation role was observed in S . aureus prophages . Although the xis genes are absent in some staphylococcal prophages , a gene structure called ORF-C may play a similar role instead [43] . The deletion of the gene locus of int in Ф11 resulted in a permanent lock of the prophage genome onto the host chromosome but not in a loss of the phage , while the excision and integration of Φ12 and Φ13 were not affected . No excision of the Ф11 genome could be detected though integrases of Φ12 and Φ13 still remained . The result reflected the high specificity of the integrases . The frequencies of the spontaneous excisions of prophage genomes varied among different S . aureus prophages . The ratio might partly depend on the sequence of attachment site , although some other factors are involved . For instance , the proportion of the naturally excised genome of ФSa2mw is 100 times of that of Ф12 even though they share the same attB site and integrase . We uncovered the fact that the host alternative sigma factor σH directly modulated the transcription of S . aureus prophage integrase . The deletion of sigH caused an obvious decrease in integrase expression . This finding was confirmed by experiments both in vitro and in vivo . σH was also found to control the expression of the comG and comE operons , which may play roles in genetic competence in S . aureus [11] . Genetic competence may be defined as a physiological state for taking up high-molecular weight exogenous DNA [44] . But currently there has been no evidence that supports any connections between the phage's life cycle and the competent state of the host . While σA is commonly utilized for promoter recognition at the immunity region , the finding of the recruitment of a second sigma factor by the phage appears especially appealing . Besides , the Northern blot assay showed that integrase could be expressed by another long transcript from a σA-dependent promoter . This result demonstrated that σH presumably affected the transcriptional level of integrase in a mildly regulative manner . The deletion of sigH would cause an increase in naturally excised phage genomes . Because the process of integration and excision is reversible , the excised forms of phage genomes may reintegrate into the host genome or occasionally enter the lytic cycle . Studies on S . aureus prophage gene arrangement showed that int genes were adjacent to the attL site and usually no ORFs were found downstream in the same direction . And Northern blot assays showed that the transcription on the opposite direction was unaffected in the sigH mutant . Therefore , we deduced that the expression of neighboring genes ( i . e . xis/ORF-C ) should not be modulated by the sigH gene . It can be postulated that the increase in spontaneous phage excision was more likely a result of a decrease in reintegration rather than an increase in excision due to the reduction of integrase . It was previously reported that lysis genes were more active in excised phage DNA molecules in streptococcal prophages [34] . Here we also found that the expression of cI was decreased in the int mutant probably because of the immobility of the prophage . Besides , the spontaneous lysis rate in the sigH mutant was higher . This phenomenon was probably a consequence of a higher excision frequency of free circular phage genomes . Notably , the transcriptional levels of accessory virulence genes ( i . e . sak , sea ) were intimately associated with the phage's life cycle as well [28] . A lower rate of lysogenization was observed in the sigH mutant , probably due to the absence of the short transcript . Although the mRNA transcribed from a σA-dependent promoter at the immunity region also expresses integrase , it is postulated that integrase expressed from the short transcript is dominant once a phage infects a bacterial cell because the σH-dependent promoter is much closer to the ORF of int . This mechanism is very similar to the one in the well-studied bacteriophage lambda . A small transcript of int is produced from pRE under the stimulation of protein CІІ , while a long transcript from pL can also express integrase in a stable lysogenic state [45] , [46] . Nevertheless , the difference between them is obvious in that CII protein is encoded by a phage gene at early stage to ensure the precise regulation [47] while sigH is a host-source protein . Although the absence of sigH does not cause an intense variation in the prophage characteristics , the presence of sigH has a distinct function in promoting and stabilizing the lysogeny of prophages in S . aureus . This may also explain why the conserved promoter region has been kept in most staphylococcal prophages during evolution but has never been found in other firmicutes harboring phages . The regulation of sigH remains largely unknown . As an alternative sigma factor , the expression level of sigH would possibly respond to certain cellular conditions and/or variations in the environment . The findings suggest that the recruitment of sigH may provide the staphylococcal temperate phages with an additional strategy to sense the host conditions and/or the change of living environments . This study could be vital in understanding the temperate phage lysogenic strategy and phage-related virulence . Further investigations would mainly concentrate on which host physiological conditions and/or their circumstance S . aureus sigH would respond to and how that process is regulated .
The bacterial strains , plasmids , and oligonucleotides used in this study are listed in Tables S1 , S2 , and S3 . Bacteria were routinely grown in Luria-Bertani ( LB ) medium ( for E . coli ) or Tryptone Soy Broth ( TSB , Oxiod ) medium ( for S . aureus ) with aeration ( 200 rpm ) at 37°C . For the antibiotics supplement in cultivation , 100 µg/ml ampicillin , 50 µg/ml kanamycin or 34 µg/ml chloromycetin was used for E . coli strains; 15 µg/ml chloromycetin or 10 µg/ml erythromycin was used for S . aureus strains . All plasmids used in S . aureus were first transferred into strain RN4220 [48] and then transformed into MW2 or NCTC8325 by electroporation . For genomic DNA or RNA extraction , one colony of each sample was inoculated in 5 ml of TSB medium and incubated at 37°C overnight . Each culture was started by diluting the precultures to an OD600 = 0 . 05 and was then incubated at 37°C ( 200 rpm ) . The cultivation was stopped at early exponential phase ( OD600 = 0 . 2 ) , mid-exponential phase ( OD600 = 0 . 6 ) , late-exponential phase ( OD600 = 2 . 4 ) , and stationary phase ( OD600 = 4 . 0 ) , respectively . S . aureus cells were pre-digested with digestion buffer containing 40 U/ml lysostaphin , 10 mg/ml lysozyme and 10% ( v/v ) glycerol . Genomic DNA was extracted using the EZ-10 Spin Column Genomic DNA Isolation Kit ( Bio Basic Inc . ) . RNA extraction was performed using the SV Total RNA Isolation System ( Promega ) . Residue DNA in extracted RNA was removed by treatment with 10 U of DNaseІ ( Takara ) at 37°C for 1 hour . RNA was purified by phenol-chloroform extraction and ethanol precipitation . Purified total RNA and genomic DNA were qualified and quantified by DU730 Nucleic Acid/Protein Analyzer ( Beckman Coulter ) for reverse transcription , Q-PCR and/or Northern blot assay . Reverse transcription was carried out following the technical manual of ImProm-ІІ Reverse Transcription System ( Promega ) . Q-PCR was performed using StepOne Real-time System ( Applied Biosystems ) . Northern blot assays were performed by using the BrightStar BioDetect Kit ( Ambion ) . Probes for Northern blotting were created by using an Ambion BrightStar Psoralen-Biotin Nonisotopic Labeling Kit . The promoter region of Φ11 integrase was amplified by PCR with primers pint11-F and pint11-R . The DNA ladder was created using ABI PRISM BigDye Terminators kit with primer pint11-R . 5′-FAM-labelled primer pint11-R-EX was purchased from Sangon ( Shanghai , China ) . For primer extension assay , 50 µg of total RNA of S . aureus NCTC8325 from the exponential phase was used . The reverse transcription was performed with pint11-R-EX at 42°C for 1 hour according to the manufacture's instructions of Primer Extension System AMV Reverse Transcriptase ( Promega ) . The product of reverse transcription and DNA ladder were then separated by capillary electrophoresis and fluorescent signals were collected by ABI3770 sequencer ( Applied Biosystems ) . The sigH and int genes of Φ11 in S . aureus RN4220 , MW2 and/or NCTC8325 were knocked out using a temperature sensitive shuttle vector pMAD [49] . DNA sequences of about 600 bps , located at the up- and down-stream of ORFs of target genes , were amplified by PCR and inserted into pMAD consecutively to obtain pMADΔsigH and pMADΔint11 . The constructed plasmids were first transformed into S . aureus strain RN4220 and later transferred into MW2 or NCTC8325 by electroporation . Mutant strains were obtained by a two-step screen method as previously described [49] . In the mutants , only the ORFs of target genes were deleted and no extra genes ( i . e . antibiotics resistance genes ) were introduced into the bacterial genomes . To obtain sigH and sigA proteins for assays in vitro , sigH and sigA genes were amplified from the S . aureus NCTC8325 genome by PCR and inserted into pET22b at the site of NdeІ/XholІ to generate pET22bsigA and pET22bsigH . The plasmids were transformed into the E . coli strain Rosetta ( DE3 ) to express his-tag fusion proteins under the induction of 1 mM IPTG at 16°C . Target proteins were purified from cell lysate by Ni-NTA resin , eluted and then dialyzed to remove imidazole . The promoter regions of cro and int of Φ11 were amplified with primers pcro11-F , pcro11-R , pint11-F and pint11-R by PCR and purified using QIAquick PCR Purification Kit ( Qiagen ) to obtain the DNA templates containing pcro or pint . One microgram of E . coli core RNA polymerase ( Epicentre ) was pre-incubated with or without sigA or sigH ( 400 ng ) at 37°C for 5 min . Then , 5 µl of DNA template ( 1 µg ) , 1 µl of NTP mixture ( 2 . 5 mM each ) , 10 µCi of [α-32P]UTP ( 5000 Ci/mM ) and 40 U of RNase inhibitor ( Takara ) were added to form a final volume of 50 µl containing 0 . 04 M Tris-Cl ( pH 7 . 5 ) , 0 . 15 M KCl , 10 mM MgCl2 , 0 . 01% Triton X-100 and 0 . 02 M dithiothreitol . The mixture was incubated at 37°C for 20 minutes , and the reaction was stopped by 0 . 2 M sodium dodecyl sulfate . The reaction products were purified by phenol-chloroform extraction and then ethanol precipitation . The pellet was resuspended in formamide and denatured at 95°C for 5 minutes . Samples were then separated on an 8% polyacrylamide gel with 6 M urea for radioautography . To detect the σH-pint recognition in vivo , a conditional replication and integration plasmid pAH125 was deployed [29] . The promoter region of Φ11 int including SD sequence was amplified with the primers pint11-AH125-F and pint11-AH125-R by PCR . The amplicon was cloned into multiple cloning site of pAH125 at the site of PstІ/EcoRІ to create pint-lacZ fusion . Construction of pAH125pint was operated in the E . coli strain BW25142 . Plasmid pAH125pint was transformed into the E . coli strain ZK126 and then integrated into the host genome with the help of pINTts to gain strain TW0901 as previously described [29] . The T7 promoter of pET22bsigH was replaced by tac promoter with lac operator from pGEX-2T at BglІІ/NdeІ site to generate pET22btac-sigH . Strain TW0902 was later obtained by introducing pET22btac-sigH into TW0902 . Overnight grown cultures of ZK126 , TW0901 , and TW0902 were diluted into 50 ml of LB medium , respectively , with an OD600 = 0 . 05 to start the incubation at 37°C ( 200 rpm ) . After OD600 reached 0 . 7 , the cultures were transferred to 16°C , with or without 1 mM IPTG induction , for an additional 4 hours . The harvest cultures were centrifuged to remove the liquid medium and resuspended in phosphate buffer containing 60 mM Na2HPO4 , 40 mM NaH2PO4 ( pH 7 . 0 ) , 10 mM KCl , 1 mM MgSO4 and 50 mM β-mercaptoethanol to an OD600 = 1 . 0 for lysis . The activity of β-galactosidase was determined as previously described [50] . Phage induction by UV treatment was performed as previously described [51] . One milliliter of culture at exponential phage ( OD600 = 0 . 6 ) was diluted in 9 ml of chilled phosphate buffered saline ( PBS , pH 7 . 4 ) . The mixture was irradiated with a dose of 50 J/m2 on 9 cm diameter Petri dishes . Two milliliters of 5×TBS was immediately added after the irradiation . The culture was then incubated at 37°C in the dark for an additional 2 hours . Phage titers were determined by the traditional double layer method [52] using RN4220 as the indicator strain . To obtain the phage lysate from lysogens , 2 mg/ml mitomycin-C was added into the cultures at exponential phase , followed by further incubation for 4 hours [53] . Supernatants were sterilized using 0 . 45 pore diameter membrane filters ( Millipore ) . Phage particles were concentrated by precipitation with polyethylene glycol as previously described [54] . The phage lysate from NCTC8325 by treatment with mitomycin-C was spotted on the susceptible strain RN4220 . Survival clones from the center of the plaques were picked and checked for lysogeny by PCR . RN4220Φ11 , RN4220Φ12 , and RN4220Ф13 were obtained for subsequent experiments by using this method . To measure the lysogenization ability of the S . aureus strains , Φ11 lysate with a titer of about 5×108 was prepared . Susceptible cells were mixed with Φ11 particles with a multiplicity of 50 in PBS containing 10 mM Mg2+ at 37°C for 4 hours . The supernatants were then discarded to remove the free viral particles . Pellets were resuspended in PBS and plated onto agar plates . The survival clones were checked for lysogeny by cross-streak assays , which were performed as follows . The bacteriophage lysate was applied as a narrow band across the center of an agar plate , and bacteria were streaked across the dried lysate . After incubation at 37°C for 12 hours , a clear spot was visualized at the cross if the tested bacteria were susceptive cells but not the lysogens . Genbank accession number of S . aureus NCTC8325: NC_007795 . Genbank accession number of S . aureus MW2: NC_003923 . Genbank available phage genomes: NC_004615 ( Φ11 ) , NC_004616 ( Φ12 ) , NC_004617 ( Φ13 ) , NC_007047 ( Φ187 ) , NC_007051 ( Φ2638A ) , NC_007055 ( Φ37 ) , NC_007053 ( Φ3A ) , NC_007052 ( Φ42E ) , NC_007060 ( Φ55 ) , NC_007049 ( Φ53 ) , NC_007062 ( Φ52A ) , NC_007061 ( Φ29 ) , NC_007048 ( Φ69 ) , NC_007059 ( Φ71 ) , NC_007050 ( Φ85 ) , NC_007063 ( Φ88 ) , NC_7064 ( Φ92 ) , NC_7057 ( Φ96 ) , NC_007065 ( ΦX2 ) , NC_005356 ( Φ77 ) , NC_009526 ( Φ80α ) , NC_007056 ( ΦEW ) , NC_002951 ( ΦCOL ) , NC_003288 ( ΦETA ) , NC_008798 ( ΦETA2 ) , NC_008799 ( ΦETA3 ) , NC_010147 ( ΦMR11 ) , NC_010808 ( ΦMR25 ) , NC_004740 ( ΦN315 ) , NC_008583 ( ΦNM1 ) , NC_008617 ( ΦNM3 ) , NC_008689 ( ΦPVL108 ) , NC_011612 ( ΦSauS-IPLA35 ) , NC_011614 ( ΦSauS-IPLA88 ) , NC_002661 ( ΦSTL ) , NC_002321 ( ΦPVL ) , NC_002486 ( ΦPV83 ) , NC_007058 ( ΦROSA ) , NC_007055 ( Φ37 ) , NC_008722 ( ΦCNPH82 ) , NC_008723 ( ΦPH15 ) , NC_007168 ( ΦSH1 ) , NC_007457 ( ΦCherry ) , NC_009815 ( ΦA006 ) , NC_001884 ( ΦSPBc2 ) , NC_005822 ( ΦLC3 ) , NC_005294 ( ΦEJ-1 ) , NC_009819 ( ΦP9 ) , NC_003524 ( Φ3626 ) , NC_009231 ( ΦC2 ) . Genbank GeneID of S . aureus NCTC8325 sigH: 3920368 . Genbank GeneID of S . aureus MW2 sigH: 1002599 . Genbank GeneID of S . epidermidis RP62A sigW: 3241810 . Genbank GeneID of S . haemolyticus JCSC1435 sigH: 3482668 . Genbank GeneID of B . anthracis Sterne sigH: 2851339 . Genbank GeneID of B . subtilis 168 sigH: 936150 . Genbank GeneID of Clostridium difficile 630 sigH: 4916669 . Genbank GeneID of Clostridium perfringens 13 sigH: 990784 . Genbank GeneID of Listeria monocytogenes Clip81459 sigH: 7703665 . Genbank GeneID of S . aureus NCTC8325 Φ11 int: 1258054 . | Staphylococcus aureus is a widely distributed opportunistic pathogen causing numerous foreign-body-associated infections . A large group of virulence factors are encoded by genes of prophages integrated in the bacterial genome . Here we show a heretofore unrecognized mechanism whereby an alternative sigma factor is recruited by a staphylococcal temperate phage for the regulation of int transcription . The modulation is processed via a direct recognition of a newly defined int promoter , while the integrase has critical roles in prophage integration and excision . The recruitment of a host-source sigma factor for integration modulation may provide the prophage with a novel strategy to sense the host conditions and further influence prophage gene expression and correlative bacterial virulence . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbiology/parasitology",
"microbiology/medical",
"microbiology",
"microbiology"
] | 2010 | Alternative Sigma Factor σH Modulates Prophage Integration and Excision in Staphylococcus aureus |
Patients with neonatal severe hyperparathyroidism ( NSHPT ) are homozygous for the calcium-sensing receptor ( CaR ) mutation and have very high circulating PTH , abundant parathyroid hyperplasia , and severe life-threatening hypercalcemia . Mice with homozygous deletion of CaR mimic the syndrome of NSHPT . To determine effects of CaR deficiency on skeletal development and interactions between CaR and 1 , 25 ( OH ) 2D3 or PTH on calcium and skeletal homeostasis , we compared the skeletal phenotypes of homozygous CaR–deficient ( CaR−/− ) mice to those of double homozygous CaR– and 1α ( OH ) ase–deficient [CaR−/−1α ( OH ) ase−/−] mice or those of double homozygous CaR– and PTH–deficient [CaR−/−PTH−/−] mice at 2 weeks of age . Compared to wild-type littermates , CaR−/− mice had hypercalcemia , hypophosphatemia , hyperparathyroidism , and severe skeletal growth retardation . Chondrocyte proliferation and PTHrP expression in growth plates were reduced significantly , whereas trabecular volume , osteoblast number , osteocalcin-positive areas , expression of the ALP , type I collagen , osteocalcin genes , and serum ALP levels were increased significantly . Deletion of 1α ( OH ) ase in CaR−/− mice resulted in a longer lifespan , normocalcemia , lower serum phosphorus , greater elevation in PTH , slight improvement in skeletal growth with increased chondrocyte proliferation and PTHrP expression , and further increases in indices of osteoblastic bone formation . Deletion of PTH in CaR−/− mice resulted in rescue of early lethality , normocalcemia , increased serum phosphorus , undetectable serum PTH , normalization in skeletal growth with normal chondrocyte proliferation and enhanced PTHrP expression , and dramatic decreases in indices of osteoblastic bone formation . Our results indicate that reductions in hypercalcemia play a critical role in preventing the early lethality of CaR−/− mice and that defects in endochondral bone formation in CaR−/− mice result from effects of the marked elevation in serum calcium concentration and the decreases in serum phosphorus concentration and skeletal PTHrP levels , whereas the increased osteoblastic bone formation results from direct effects of PTH .
The extracellular calcium-sensing receptor ( CaR ) is a G protein-coupled receptor that plays an essential role in the regulation of extracellular calcium homeostasis . This receptor is expressed in nearly all tissues [1] . Cloning of the CaR was immediately followed by the association of human genetic diseases with inactivating or activating CaR mutations: familial hypocalciuric hypercalcemia ( FHH ) and neonatal severe hyperparathyroidism ( NSHPT ) are caused by CaR-inactivating mutations , whereas autosomal dominant hypoparathyroidism is secondary to CaR-activating mutations [1] . Patients with FHH are heterozygous for the CaR mutation and have a normal or mildly increased circulating parathyroid hormone ( PTH ) level , normal parathyroid histology or mild parathyroid hyperplasia , and mild to moderate hypercalcemia [2] , [3] . Patients with NSHPT are homozygous for the CaR mutation and have very high circulating PTH , abundant parathyroid hyperplasia and severe life-threatening hypercalcemia [4] . Targeted inactivation of the CaR gene in mice has resulted in the development of models of the human syndromes [5] . Thus mice with heterozygous deletion of CaR mimic the syndrome of FHH , whereas homozygotes mimic the syndrome of NSHPT and generally die within a few days to weeks after birth . A wide variety of functions have been attributed to CaR . Previous studies confirmed that most CaR−/− mice died within 2 weeks after birth [6] . Compared to WT littermates , body size and body weight were reduced markedly , serum calcium levels were markedly elevated , serum phosphorus levels were decreased , serum PTH levels were raised significantly in 2-week-old CaR−/−mice , and the parathyroid glands were also enlarged . CaR−/− mice die shortly after birth because of the effects of severe hyperparathyroidism and hypercalcemia . However , the lethal CaR–deficient phenotype has made it difficult to dissect direct effects of CaR deficiency from secondary effects of hyperparathyroidism and hypercalcemia . Recently , targeted deletion of the CaR from chondrocytes has been reported to be lethal in utero in mice before embryonic day 13 but to produce viable mice with delayed growth plate development if conditional targeted deletion in these cells is induced between E16 and E18 [7] . Targeted deletion of the CaR from early committed osteoblasts resulted in smaller , undermineralized skeletons with significant reductions in bone volume and bone mineral density in the femur and vertebrae [7] . This study demonstrated a critical role for the CaR in skeletal development . However , previous studies failed to find such role for CaR in mice with targeted disruption of exon 5 of the CaR gene encoding a portion of the extracellular domain of this receptor [5] , [6] . It has not been established that CaR has a direct role in the skeleton based on this animal model . As noted earlier , these observations may be confounded by the severe hyperparathyroidism and the accompanying hypercalcemia and hypophosphatemia in this animal model . To better understand direct effects of CaR on bone and cartilage function , correction of hyperparathyroidism is required in this CaR–deficient mouse model . We have previously reported a mouse model deficient in PTH by targeting the Pth gene in embryonic stem cells [8] , [9] . Although adult Pth-null mice develop hypocalcemia , hyperphosphatemia and low circulating 1 , 25 ( OH ) 2D3 levels consistent with primary hypoparathyroidism [9] , this phenotype is not lethal . Therefore , a double-knockout model was established by crossing the PTH-deficient with the CaR–deficient mice [10] to correct the severe hyperparathyroidism , hypercalcemia and hypophosphatemia observed in the homozygous CaR–deficient mice . The results of this study indicated that elimination of hyperparathyroidism rescued the increased neonatal mortality as well as the rickets-like skeletal abnormality in these mice suggesting that the early lethality could be corrected by eliminating the hypercalcemia and hyperparathyroidism [10] , [11] . However , any essential , nonredundant role for CaR in regulating chondrogenesis or osteogenesis could not be identified based on analysis of the skeleton of this adult double knockout model . It is unclear why there are discrepancies between the findings from CaR conditional knockout mice [7] and those from CaR conventional knockout mice [6] , [10] regarding the physiological significance of the CaR in the skeleton , although residual biological activity of an alternatively spliced CaR lacking exon 5 in the conventional , global knockout model may be one explanation . The vitamin D-PTH axis plays a central role in calcium and phosphate homeostasis and is essential for skeletal development and mineralization . PTH and 1 , 25-dihydroxyvitamin D3 [1 , 25 ( OH ) 2D3] directly affect calcium homeostasis and each exerts important regulatory effects on the other . PTH stimulates the production of 1 , 25 ( OH ) 2D3 by activating the renal 25-hydroxyvitamin D-1α-hydroxylase [1α ( OH ) ase] [12] , [13]; 1 , 25 ( OH ) 2D3 , in turn , suppresses the production of PTH [14] , [15] and controls parathyroid cell growth [16] . 1 , 25 ( OH ) 2D3 suppression of PTH synthesis occurs through negative regulation of PTH gene transcription by a 1 , 25 ( OH ) 2D3-vitamin D receptor ( VDR ) /retinoid X receptor ( RXR ) complex [17] in the parathyroid cells [18] . We [19] and others [20] have previously reported a mouse model deficient in 1 , 25 ( OH ) 2D by targeted ablation of the 1α ( OH ) ase gene [1α ( OH ) ase−/−] . After being weaned , mice fed a diet of regular mouse chow developed secondary hyperparathyroidism , retarded growth and skeletal abnormalities characteristic of rickets . These abnormalities mimic those described in vitamin D-dependent rickets type I [21] and this mouse phenotype is not lethal . Furthermore , by comparing mice with targeted disruption of the PTH or 1α ( OH ) ase genes with PTH-1α ( OH ) ase double null mice , we found that PTH−/−1α ( OH ) ase−/− mice died of tetany with severe hypocalcemia by 3 weeks of age with severe defects in skeletal development [22] . To determine effects of CaR deficiency on skeletal development as well as to investigate interactions between CaR , 1 , 25 ( OH ) 2D3 and PTH on calcium and skeletal homeostasis , we compared the skeletal phenotypes of CaR−/− mice to those of the double homozygous , CaR- and 1α ( OH ) ase-deficient [CaR−/−1α ( OH ) ase−/−] mice or those of the double homozygous CaR- and PTH-deficient [CaR−/−PTH−/−] mice at 2 weeks of age .
In CaR−/− mice , the survival rate was 20% and the body weight was decreased significantly at 2 weeks of age; Ablation of 1α ( OH ) ase in CaR−/− mice resulted in a slightly longer lifespan and a significant increase of body weight; Ablation of PTH in CaR−/− mice resulted in a normalized lifespan and body weight ( Figure 1A–1D ) . At 2 weeks of age , CaR−/− mice displayed hypercalcemia , hypophosphatemia , and increased serum ALP , PTH and 1 , 25 ( OH ) 2D3; 1α ( OH ) ase−/− mice displayed mild hypocalcemia , hypophosphatemia , hyperparathyroidism , and undetectable serum 1 , 25 ( OH ) 2D3; PTH−/− mice had moderate hypocalcemia and hyperphosphatemia , lower serum 1 , 25 ( OH ) 2D3 and undetectable serum PTH; Ablation of 1α ( OH ) ase in CaR−/− mice resulted in normocalcemia , more severe hypophosphatemia , and more marked elevations in serum ALP and PTH; In contrast , ablation of PTH in CaR−/− mice resulted in normocalcemia , less severe hyperphosphatemia , normal serum ALP levels , lower serum 1 , 25 ( OH ) 2D3 and undetectable serum PTH ( Table 1 ) . To determine potential interactions between the effects of CaR and 1 , 25 ( OH ) 2D3 or PTH on parathyroid gland growth , the size of the parathyroid glands were examined by histology and image analysis . The parathyroid glands were mildly enlarged in CaR−/− mice ( Figure 1E–1H ) and 1α ( OH ) ase−/− mice ( Figure 1E and 1G ) , and were significantly enlarged in PTH−/− mice compared ( Figure 1F and 1H ) to the wild-type mice . Ablation of 1α ( OH ) ase in CaR−/− mice resulted in further enlargement of the parathyroid glands ( Figure 1E and 1G ) . Ablation of PTH in CaR−/− mice also resulted in more marked enlargement of parathyroid glands , which , however , were only slightly more enlarged relative to PTH−/− mice ( Figure 1F and 1H ) . To assess the interaction between CaR and 1 , 25 ( OH ) 2D3 or PTH on skeletal development , skeletal phenotypes of gender-matched wild-type and the five mutant models were analyzed at 2 weeks of age by radiography , micro-CT and histology . Radiographs of femurs demonstrated that the lengths of femurs were dramatically reduced in CaR−/− mice , while those of the 1α ( OH ) ase−/− and PTH−/− mice were only slightly shortened compared with their wild-type littermates ( Figure 2A ) . Ablation of 1α ( OH ) ase in CaR−/− mice resulted in an increase in the lengths of femurs , but they were still shorter than in wild-type mice , whereas ablation of PTH in CaR−/− mice resulted in normalization of the lengths of femurs ( Figure 2A ) . Radiolucency in metaphyses and diaphyses was increased markedly and no mineralized epiphyses were detected in CaR−/− mice . Radiolucency in whole femurs was slightly increased in 1α ( OH ) ase−/− and PTH−/− mice . Ablation of 1α ( OH ) ase in CaR−/− mice reduced the radiolucency , especially in the metaphyses , and ablation of PTH in CaR−/− mice also reduced the radiolucency relative to CaR−/− mice; however , the radiolucency was still greater than that in their wild-type littermates ( Figure 2A ) . From longitudinal and cross sections of three-dimensional reconstructed proximal ends of tibiae and middle diaphyses ( Figure 2B ) , it can been seen that mineralized trabecular bone volume in metaphyses was increased in CaR−/− mice , although no mineralized epiphyses were detected , and mineralized cortical bone volume was reduced in these mice . Mineralized epiphyseal volume , cortical and trabecular volume were reduced and the width of unmineralized growth plates was greater in 1α ( OH ) ase−/− mice . Mineralized cortical and trabecular bone volumes were reduced in PTH−/− mice . Ablation of 1α ( OH ) ase in CaR−/− mice resulted in a marked increase in mineralized trabecular bone volume , whereas ablation of PTH in CaR−/− mice nearly normalized mineralization in epiphyses , although mineralized cortical and trabecular bone volumes were still lower than in their wild-type littermates ( Figure 2B ) . Longitudinal sections of proximal ends of tibiae stained histochemically for total collagen are shown in Figure 2C and 2E , and trabecular volume was measured ( Figure 2G and 2I ) . Trabecular bone volume was increased significantly in CaR−/− mice and reduced in 1α ( OH ) ase−/− and PTH−/− mice relative to their wild-type littermates . Ablation of 1α ( OH ) ase in CaR−/− mice resulted in a more dramatic increase in the trabecular bone volume , whereas ablation of PTH in CaR−/− mice resulted in a marked reduction in the trabecular bone volume , even compared to PTH−/− mice ( Figure 2C , 2E , 2G and 2I ) . Osteoid volume was increased significantly in CaR−/− and CaR−/−1α ( OH ) ase−/− mice , and was not altered significantly in 1α ( OH ) ase−/− , PTH−/− and CaR−/−PTH−/− mice relative to their wild-type littermates ( Figure 2D , 2F , 2H and 2J ) . To assess the interaction between CaR and 1 , 25 ( OH ) 2D3 or PTH on endochondral bone formation , the phenotypes of growth plates were analyzed at 2-weeks of age by histology and immunohistochemistry . Only a few hypertrophic chondrocytes were observed in secondary ossification centers , and the width of growth plates and hypertrophic zones were increased dramatically in CaR−/− mice , increased in 1α ( OH ) ase−/− mice and not altered in PTH−/− mice relative to their wild-type littermates . Ablation of 1α ( OH ) ase in CaR−/− mice accelerated slightly the formation of secondary ossification centers , resulted in more hypertrophic chondrocytes and the appearance of some osteoblasts and the width of growth plates and hypertrophic zones were reduced significantly . Ablation of PTH in CaR−/− mice normalized the secondary ossification centers and the growth plates ( Figure 3A , 3D , 3E and Figure 4A , 4D , 4E ) . The proliferation of chondrocytes was assessed by immunostaining for proliferating cell nuclear antigen ( PCNA ) , and the percentage of PCNA-positive chondrocytes was quantitated . The percentage of PCNA-positive chondrocytes was reduced dramatically in CaR−/− mice , and was not altered significantly in 1α ( OH ) ase−/− and PTH−/− mice . Ablation of 1α ( OH ) ase in CaR−/− mice increased significantly the percentage of PCNA-positive chondrocytes; however , it was still lower than in their wild-type littermates . Ablation of PTH in CaR−/− mice increased the percentage of PCNA-positive chondrocytes more dramatically , to a level even higher than that in their wild-type littermates ( Figure 3B , 3F and Figure 4B , 4F ) . In view of the fact that PTHrP plays an important role in the proliferation and apoptosis of chondrocytes , we examined the localization of PTHrP in chondrocytes by immunostaining and the expression of the PTHrP gene in cartilaginous growth plates . The percentage of PTHrP-positive chondrocytes was reduced markedly in CaR−/− mice , reduced significantly in 1α ( OH ) ase−/− mice and increased in PTH−/− mice . Ablation of 1α ( OH ) ase in CaR−/− mice increased significantly the percentage of PTHrP-positive chondrocytes; however , it was still lower than in their wild-type littermates . Ablation of PTH in CaR−/− mice increased markedly the percentage of PTHrP-positive chondrocytes ( Figure 3C , 3G and Figure 4C , 4G ) . Alteration in the pattern of expression of the PTHrP gene in cartilaginous growth plates were similar to those observed for the percentage of PTHrP positive chondrocytes ( Figure 3H and Figure 4H ) . To determine whether alterations in trabecular bone volume were associated with those in osteoblastic bone formation , paraffin sections from the various genotypes of mice at 2 weeks of age were stained with HE ( Figure 5A and 5C ) as well as immunohistochemically for osteocalcin ( Figure 5B and 5D ) . The number of osteoblasts ( Figure 5E and 5F ) and osteocalcin-positive areas ( Figure 5G and 5H ) in metaphyseal regions were quantitated by image analysis . The number of osteoblasts and osteocalcin-positive areas were increased significantly in CaR−/− mice , while they were reduced in 1α ( OH ) ase−/− and PTH−/− mice relative to their wild-type littermates . Ablation of 1α ( OH ) ase in CaR−/− mice resulted in a more dramatic increase in the number of osteoblasts and osteocalcin-positive areas , whereas ablation of PTH in CaR−/− mice resulted in a marked reduction in the number of osteoblasts and osteocalcin-positive areas , even compared to PTH−/− mice ( Figure 5A–5H ) . We also examined alterations in the expression of genes related to bone formation . RNA was isolated from long bones and real-time RT–PCR was performed . Results showed that the expression of the ALP , type I collagen and osteocalcin genes were increased significantly in CaR−/− mice , while they were reduced in 1α ( OH ) ase−/− and PTH−/− mice relative to their wild-type littermates . Ablation of 1α ( OH ) ase in CaR−/− mice resulted in a more dramatic increase in the expression of these same genes , whereas ablation of PTH in CaR−/− mice resulted in a marked reduction in their expression , even compared to PTH−/− mice ( Figure 5I and 5J ) . These alterations were consistent with those of osteoblastic bone formation parameters observed by histomorphometric analysis . To assess the interaction between CaR and 1 , 25 ( OH ) 2D3 or PTH on osteoclastic bone resorption , paraffin sections were stained histochemically for TRAP ( Figure 6A and 6B ) . Osteoclast number ( Figure 6C and 6D ) and surface ( Figure 6E and 6F ) were determined by image analysis . The results revealed that the TRAP-positive osteoclast number and surface were increased markedly in the metaphyseal regions in CaR−/− mice and were decreased significantly in 1α ( OH ) ase−/− and PTH−/− mice relative to their wild-type littermates . Ablation of 1α ( OH ) ase in CaR−/− mice decreased significantly osteoclast number and surface; however , these parameters were still greater than in their wild-type littermates . Ablation of PTH in CaR−/− mice markedly decreased osteoclast number and surface ( Figure 6A–6F ) . We also examined alterations in the expression of genes related to bone resorption . RNA was isolated from long bones . The gene expression of the RANKL and OPG genes were examined by real-time RT–PCR , and the ratio of RANKL/OPG mRNA levels was calculated . Results revealed that alterations in the ratio of RANKL/OPG mRNA levels were consistent with those of TRAP-positive osteoclast number and surface ( Figure 6G and 6H ) .
In the present study , we examined the effect on mineral homeostasis of deleting CaR ( CaR−/− ) alone , or both 1 , 25 ( OH ) 2D3 and CaR ( 1α ( OH ) ase−/−CaR−/− ) , or both PTH and CaR ( PTH−/−CaR−/− ) . Our results confirmed previous findings in CaR−/− mice: namely the presence of hypercalcemia , hypophosphatemia , hyperparathyroidism and enlarged parathyroid glands [5] , [6] . Deletion of 1α ( OH ) ase in the CaR null background corrects the hypercalcemia but leads to more severe hypophosphatemia , hyperparathyroidism and much larger parathyroid glands . These alterations were associated with diminished capacity of double knockout mice to stimulate absorption of calcium and phosphate from the gastrointestinal tract and to repress PTH biosynthesis , and resulted in a further stimulus for parathyroid gland enlargement , overproduction of parathyroid hormone and consequent effects on impeding renal phosphate reabsorption . Interestingly , deletion of PTH in CaR−/− mice also corrects the hypercalcemia , and the mice develop hyperphosphatemia and undetectable serum PTH with enlarged parathyroid glands . These alterations were associated with diminished capacity of double mutant animals to stimulate calcium reabsorption and to inhibit phosphate reabsorption from the renal tubules [23] . Reduction of hypercalcemia in the CaR−/− mice was the common biochemical alteration induced by deleting the 1α ( OH ) ase in the CaR−/−1α ( OH ) ase−/− mice and PTH in the CaR−/−PTH−/− mice . Improvement in hypercalcemia is therefore likely to have been the major contributor to the improved longevity of the CaR−/− mice . Previous studies have demonstrated that 1 , 25 ( OH ) 2D3 acts via the nuclear vitamin D receptor ( VDR ) to bind to a negative response element tentatively identified in the PTH promoter , which can repress PTH gene expression [17] . We previously showed that 1 , 25 ( OH ) 2D3 can inhibit the growth of primary cultures of bovine parathyroid cells in vitro [24] . We also showed that 1 , 25 ( OH ) 2D3 inhibits parathyroid hyperplasia in vivo in a mouse model with targeted ablation of the 1α ( OH ) ase gene [25] . Our current study found more severe hypophosphatemia and hyperparathyroidism with enlarged parathyroid glands in 2-week-old double homozygous CaR- and 1α ( OH ) ase-deficient neonates than in CaR−/− mice , indicating that 1 , 25 ( OH ) 2D3 inhibits parathyroid growth and PTH biosynthesis in a CaR-independent manner . We next examined the effect of CaR deficiency and the interaction between CaR and 1 , 25 ( OH ) 2D3 or PTH on endochondral bone formation . Our results show that CaR deficiency resulted in a severe delay of secondary ossification center formation , an enlarged growth plate and a significant reduction of chondrocyte proliferation and apoptosis . Genetic deficiencies in 1 , 25 ( OH ) 2D3 caused by loss of a functional 1α ( OH ) ase enzyme or vitamin D receptor ( VDR ) result in hypocalcemia , hyperparathyroidism , hypophosphatemia , rickets , and osteomalacia . These effects , however , are very mild or absent in suckling mice but become markedly accelerated in mice after weaning , when dietary calcium and lactose content decreases . The phenotype can largely be corrected by administration of a high calcium , high lactose-containing diet that bypasses active , 1 , 25 ( OH ) 2D3-mediated , Ca2+ absorption [25] . It is noteworthy that analysis of the chondrocyte-specific Cyp27b1-knockout and Cyp27b1-overexpressing mice revealed a fetal bone phenotype which did not persist however beyond the immediate neonatal period [26] . Consequently although 1 , 25 ( OH ) 2D3 may play a direct role in normal development of the cartilaginous growth plate , other factors , such as endocrine maintenance of calcium and phosphate balance may be more important in defining postnatal bone development . The deletion of 1α ( OH ) ase in CaR−/− mice slightly accelerated secondary ossification center formation , and improved chondrocyte proliferation , while deletion of PTH in CaR−/− mice rescued these abnormalities of cartilaginous development in CaR−/− mice . This result is consistent with a previous report in which correction of severe hyperparathyroidism in CaR−/− mice resulted in healing of the rickets and osteomalacia [11] . In our studies , the normalization of serum calcium in both double mutants may have accelerated the improvements in the secondary ossification centre and in the cartilaginous growth plate . Previous studies have reported that the CaR modulates growth plate chondrocyte differentiation in vitro [27]–[29] , and chondrocyte-specific CaR deletion leads to early embryonic death ( before E13 ) . However , if the deletion occurs at the later time points , such as between E16 and E18 , viable pups can be obtained but with delayed growth plate development [7] . Currently , there is no fully accepted explanation for these discrepancies in phenotype between chondrocyte-specific CaR knockout mice obtained using mice with a floxed exon 7 of the CaR gene [7] and the CaR conventional knockout mice in which exon 5 has been disrupted [6] , [10] . Because calcium regulation is so fundamental for the maintenance of normal cell function , it is plausible that cells have some other pathways that can compensate for neonatal defects resulting from deficiency of CaR . One such compensatory mechanism may be the expression of a CaR splice variant lacking exon 5 in chondrocytes from the CaR–deficient mice [30] . In the PTH−/−CaR−/− mice complete rescue of the alterations in the secondary ossification centre and in the cartilaginous growth plate were observed . In view of the fact that PTH is not synthesized in cartilage , it cannot access this avascular tissue . It seems likely that , since phosphate per se plays a critical role in growth plate development ( 30 ) , and mineralization [31] and phosphate levels are also markedly decreased in CaR−/− mice and increased by deletion of PTH , hyperphosphatemia in the CaR−/−PTH−/− mice likely contributed to the improvement in the growth plate . PTHrP-deficient mice display reduced chondrocyte proliferation , accelerated differentiation , and increased apoptosis [31] , [32] . Our previous study demonstrated that 4-month old PTH−/− mice have significantly increased serum PTHrP concentrations and PTHrP mRNA and protein levels in bone tissues , suggesting that PTHrP is required for the increased trabecular bone volume observed in adult PTH−/− mice [33] . Based on these findings , we hypothesized that altered PTHrP levels might account for the skeletal defects of CaR−/− mice and play a compensatory role in PTH/CaR double knockout mice . To test this hypothesis , we examined the levels of expression of PTHrP mRNA and protein in cartilaginous growth plates by real-time RT-PCR and immunohistochemistry , respectively . PTHrP-immunopositive chondrocytes and PTHrP mRNA levels were markedly decreased in CaR−/− mice , slightly increased in CaR−/−1α ( OH ) ase−/− mice , and increased significantly in CaR−/−PTH−/− mice compared to their wild-type littermates . These results demonstrate that down-regulation of PTHrP expression in chondrocytes is associated with defects in the growth plate caused by CaR deficiency and that up-regulation of PTHrP expression is associated with either improvement or rescue of chondrocyte abnormalities resulting from the deletion of either 1α ( OH ) ase or PTH , in the CaR–deficient mice . 1 , 25 ( OH ) 2D3 has been reported to dampen PTHrP upregulation at both the mRNA and protein levels in prostate cancer cells [34] , [35] . 1 , 25 ( OH ) 2D3 deficiency may therefore have contributed to the increases of PTHrP observed in the double mutants . Taken together , these data suggest that the down-regulation of PTHrP expression in chondrocytes may contribute to defects of cartilage in the CaR–deficient neonates and that up-regulation of PTHrP expression in chondrocytes may exert a contributory role in rescuing these abnormalities . We also examined the effect of 1 , 25 ( OH ) 2D3 or PTH deficiency in CaR−/− mice on osteoblastic bone formation . CaR–deficient mice showed increased trabecular volume , osteoblast number , and osteocalcin-positive areas , as well as increased expression of the ALP , type I collagen and osteocalcin genes and higher serum ALP levels . These osteoblastic bone formation parameters increased dramatically in 1α ( OH ) ase/CaR double knockout mice , despite the fact that these parameters were slightly decreased in 1α ( OH ) ase-deficient neonates . These results are consistent with our previous findings from 2-week-old [22] and adult 1α ( OH ) ase−/− mice [19] , [25] . When the phenotype of 1α ( OH ) ase−/− mice was analyzed , we found that the skeletal phenotype was different before and after weaning . In 2-week-old 1α ( OH ) ase−/− mice , the trabecular volume and osteoblast numbers were decreased , and the osteoid volume was not increased significantly [22] . In contrast , in 4-month-old 1α ( OH ) ase−/− mice , the trabecular volume , osteoblast number and osteoid volume were all increased significantly even on a high calcium diet containing 1 . 5% calcium in the drinking water [25] . These differences were thought to result from the elevations in circulating PTH . Breast-feeding of neonatal 1α ( OH ) ase−/− pups with milk containing a higher calcium and low level of 1 , 25 ( OH ) 2D3 had less severe hypocalcemia and subsequent less elevation of serum PTH compared to adult mutant mice . Our previous studies had shown that serum PTH was increased 1 . 5-fold at 2 weeks of age [22] , but 30-fold at 4 months of age [25] , in the 1α ( OH ) ase−/− mice compared to their wild-type counterparts . If the stimulatory effects of elevated PTH on osteoblasts could not overcome the reduction in osteoblasts due to 1 , 25 ( OH ) 2D3 deficiency , osteoblastic bone formation parameters would be reduced . If , on the other hand , the osteoblast-stimulating effects of elevated PTH overcame the decreased osteoblasts resulting from 1 , 25 ( OH ) 2D3 deficiency , then osteoblastic bone formation parameters would be increased . The presence of increased numbers of osteoblasts and increased bone matrix volume in CaR−/− and CaR−/−1α ( OH ) ase−/− neonates suggests that any requirement of either CaR or/and 1 , 25 ( OH ) 2D3 for osteoblast activation is readily overcome by the osteoblast stimulating effects of elevated PTH . This observation was further supported by deletion of PTH in the CaR−/− mice , which resulted in a significant decrease in osteoblastic bone formation compared to PTH−/− mice . The difference in osteoblastic bone formation between the PTH−/− and CaR−/−PTH−/− mice and which can be accentuated in older animals [36] is consistent with a stimulatory effect of CaR on osteoblast function . Indeed , exposure of primary osteoblasts or a variety of osteoblast-like cells to high calcium or polycationic CaR agonists , such as neomycin and gadolinium , stimulate their proliferation , differentiation and mineralization [37] , [38] . Previous studies have revealed that calcium is also an important regulator of osteoclast function . Exposing osteoclasts to high extracellular calcium concentrations results in dramatic cell retraction followed by profound inhibition of bone resorption [39]–[41] . Subsequent studies found that very high calcium inhibits the bone-resorbing activity of osteoclasts by directly acting on the CaR that is present in osteoclast precursor cells [42] and mature osteoclasts [43] , [44] . Our results found that TRAP-positive osteoclast number and surface and the ratio of RANKL/OPG were increased in CaR−/− mice . Deletion of 1α ( OH ) ase in CaR−/−neonates resulted in a decrease in osteoclastic bone resorption parameters , however , these parameters were higher than in 1α ( OH ) ase neonates . Deletion of PTH in CaR−/− neonates also resulted in a decrease in these parameters . As in our previous report [22] , the current study confirms that osteoclastic bone resorption parameters were reduced in either 1α ( OH ) ase−/− or PTH−/− neonates . Taken together , these studies suggest that the CaR , 1 , 25 ( OH ) 2D3 and PTH are all required for osteoclastic bone resorption . Results from this study indicate that normocalcemia in patients with NSHPT may lengthen their lifespan , and deletion of PTH in patients with NSHPT may normalize skeletal growth and development . A previous study has reported that the use of intravenous pamidronate controlled severe hypercalcaemia in NSHPT patients prior to parathyroidectomy [45] , Consequently , they recommended the short-term use of pamidronate in neonatal severe hyperparathyroidism to treat extreme hypercalcaemia and halt hyperparathyroid-driven skeletal demineralization in preparation for parathyroidectomy . Recently , a retrospective review has been conducted for patients managed for NSHPT over the last 10 years [46] . Five patients with NSHPT , 3 females and 2 males , presented at a mean age of 18 days . All patients had a total parathyroidectomy and autotransplantation at a mean age of 65 days , with a mean follow-up of 5 . 5 years . One patient had normal parathyroid hormone and normal calcium levels 9 . 5 years after surgery without medication . One patient had normal levels without medication for 2 years then needed calcium and vitamin D supplements thereafter ( 8 . 5 years postoperatively ) . Three patients are still on calcium and vitamin D supplementation at 5 . 5 years , 3 . 5 years , and 8 months , respectively , after surgery . Consequently , they conclude that NSHPT is managed effectively with total parathyroidectomy [46] . In summary ( Table 2 ) , CaR−/− mice had a very short lifespan , decreased body weight and displayed hypercalcemia , hypophosphatemia , elevated serum ALP , PTH and 1 , 25 ( OH ) 2D3; CaR−/−1α ( OH ) ase−/− mice had slightly increased lifespan and body weight and displayed normocalcemia , hypophosphatemia , greater elevations in PTH and ALP , and undetectable serum 1 , 25 ( OH ) 2D3; CaR−/−PTH−/− mice displayed normocalcemia , hyperphosphatemia , normal ALP levels , undetectable serum PTH and lower serum 1 , 25 ( OH ) 2D3 . CaR deficiency resulted in a severe delay of secondary ossification center formation , an enlarged growth plate and a significant reduction of chondrocyte proliferation . These alterations were associated with hypophosphatemia and decreased PTHrP expression in chondrocytes . Deletion of 1α ( OH ) ase in CaR−/− mice partially rescued the cartilage phenotype associated with an increase in PTHrP expression in chondrocytes but persistent hypophosphatemia . Deletion of PTH in CaR−/− mice rescued the phenotype of cartilage associated with the up-regulation of PTHrP expression in chondrocytes and correction of hypophosphatemia . The alterations of bone formation parameters were consistent with elevated serum PTH and with the reduction of PTHrP gene expression in bony tissue in CaR−/− and CaR−/− 1α ( OH ) ase−/− mice . Deletion of PTH in CaR−/− mice resulted in a significant decrease in osteoblastic bone formation , which was not rescued completely by increased PTHrP gene expression in bony tissue . Osteoclastic bone resorption parameters were increased in CaR−/−mice , however , they were decreased in CaR−/−1α ( OH ) ase−/− mice and more dramatically in CaR−/−PTH−/− mice compared to CaR−/− littermates . The current study demonstrates that ( 1 ) hypercalcemia , contributes to the early lethality of CaR–deficient mice , ( 2 ) defects in endochondral bone formation in CaR–deficient mice result from effects of the excess elevations in calcium and the decrease in phosphorus and PTHrP levels , while ( 3 ) the increased osteoblastic bone formation results from direct effects of PTH . Our results therefore provide mechanistic insight into the improvement in longevity and normalization in skeletal growth and development observed in patients with NSHPT after total parathyroidectomy .
The derivation of the three parental strains of CaR+/− , PTH+/−mice and 1α ( OH ) ase+/− mice by homologous recombination in embryonic stem cells was previously described by Ho et al . [5] , Miao et al . [8] , [9] and Panda et al . [19] , respectively . Briefly , a neomycin resistance gene was inserted into exon III of the mouse CaR gene or into exon III of the mouse Pth gene , replacing a portion of the extracellular domain of the CaR , which senses extracellular calcium , and the entire coding sequence of the PTH gene . Lack of CaR and PTH expression were confirmed by Western blot of kidney protein membrane extracts from homozygous CaR−/− mice [5] and by immunostaining of parathyroid glands sections [8] , respectively . A neomycin resistance gene replaced exons VI , VII , and VIII of the mouse 1α ( OH ) ase gene , removing both the ligand-binding and the heme-binding domains . RT–PCR of renal RNA from homozygous 1α ( OH ) ase−/− mice confirmed the lack of 1α ( OH ) ase expression [19] . CaR+/− and 1α ( OH ) ase+/− mice were fertile and were mated to produce offspring heterozygous at both loci , which were then mated to generate CaR−/−1α ( OH ) ase−/− pups . CaR−/−PTH−/− pups were generated in the same way as above . Lines were maintained by mating CaR+/−1α ( OH ) ase+/− males and females . These mice were of mixed genetic background , with contributions from C57BL/6J/129/SvJ and 129/SvEv/BALB/c strains . Mutant mice and control littermates were maintained in a virus- and parasite-free barrier facility and exposed to a 12-h/12-h light/dark cycle , and were fed a regular rodent diet . The use of animals in this study was approved by the Institutional Animal Care and Use Committee of Nanjing Medical University ( Approval ID 2008-00518 ) . Tail fragment genomic DNA was isolated by standard phenol–chloroform extraction and isopropanol precipitation . To determine the genotype at the CaR , PTH and 1α ( OH ) ase loci , six PCR amplification reactions were conducted . To assay the presence of the wild-type CaR allele , samples were amplified with CaR forward primer ( 5′-TCT GTT CTC TTT AGG TCC TGA AAC A-3′ ) and CaR reverse primer ( 5′-TCA TTG ATG AAC AGT CTT TCT CCC T-3′ ) . To detect the presence of the null CaR allele , Neo forward primer r-Neo-2 ( 5′-TCT TGA TTC CCA CTT TGT GGT TCT A-3′ ) was used with the CaR reverse primer . The presence of the wild-type Pth allele was detected using PTH forward primer ( 5′-GAT GTC TGC AAA CAC CGT GGC TAA-3′ ) and PTH reverse primer ( 5′-TCC AAA GTT TCA TTA CAG TAG AAG-3′ ) . The null Pth allele was detected using Neo forward primer ( 5′-TCTTGATTCCCACTTTGTGGTTCTA-3′ ) and PTH reverse primer [10] . For the wild-type Cyp27b1 [1α ( OH ) ase] allele , forward primer ( 5′-AGACTGCACTCCACTCTGAG-3′ ) and reverse primer ( 5′-GTT TCC TAC ACG GAT GTC TC-3′ ) were employed . The neomycin gene was detected with primers neo-F ( 5′-ACA ACA GAC AAT CGG CTG CTC-3′ ) , and neo-R ( 5′-CCA TGG GTC ACG ACG AGA TC-3′ ) [25] . All PCR reactions were performed with 1 cycle of 95°C for 4 minutes , and 35 cycles of 94°C for 30 seconds , 55°C for 30 seconds , 72°C for 30 seconds . Serum calcium and phosphorus were determined by autoanalyzer ( Beckman Synchron 67; Beckman Instruments ) . Serum 1 , 25 ( OH ) 2D3 was measured by radioimmunoassay ( ImmunoDiagnostic Systems , Bolden , UK ) and intact PTH was measured by a two-site immunoradiometric assay ( Immutopics , San Clemente , CA , USA ) . Femurs were removed and dissected free of soft tissue . Contact radiographs were taken using a Faxitron model 805 radiographic inspection system ( Faxitron Contact , Faxitron , Germany ) ( 22 kV voltage and 4 min exposure time ) . X-Omat TL film ( Eastman Kodak Co . , Rochester , NY , USA ) was used and processed routinely . Tibias obtained from 2-week-old mice were dissected free of soft tissue , fixed overnight in 70% ethanol and analyzed by micro-CT with a SkyScan 1072 scanner and associated analysis software ( SkyScan , Antwerp , Belgium ) as described [22] . Briefly , image acquisition was performed at 100 kV and 98 mA with a 0 . 98 rotation between frames . During scanning , the samples were enclosed in tightly fitting plastic wrap to prevent movement and dehydration . Thresholding was applied to the images to segment the bone from the background . Two-dimensional images were used to generate three-dimensional renderings using the 3D Creator software supplied with the instrument . The resolution of the micro-CT images is 11 . 26 microns . Thyroparathyroid tissue and tibias were removed and fixed in PLP fixative ( 2% paraformaldehyde containing 0 . 075 M lysine and 0 . 01 M sodium periodate ) overnight at 4°C and processed histologically as described [47] . Tibias were decalcified in ethylene- diamine tetraacetic acid ( EDTA ) -glycerol solution for 5–7 days at 4°C . Decalcified bones and other tissues were dehydrated and embedded in paraffin after which 5 µm sections were cut on a rotary microtome . The sections were stained with hematoxylin and eosin ( H&E ) or histochemically for tartrate-resistant acid phosphatase ( TRAP ) activity and for total collagen or immunohistochemical staining for osteocalcin ( OCN ) , parathyroid hormone-related protein ( PTHrP ) and proliferating cell nuclear antigen ( PCNA ) as described subsequently . Alternatively , undecalcified tibiae were embedded in LR White acrylic resin ( London Resin Company Ltd . , London , UK ) and 1-µm sections were cut on an ultramicrotome . These sections were stained for mineral with the von Kossa staining procedure . Histomorphometric indices were determined as suggested by the ASBMR Histomorphometry Nomenclature Committee [48] . Total collagen was detected in paraffin sections using a modified method of Lopez-De Leon and Rojkind [25] . Dewaxed sections were exposed to 1% sirius red in saturated picric acid for 1 h . After washing with distilled water , the sections were counterstained with hematoxylin and mounted with Biomount medium . Enzyme histochemistry for TRAP staining was performed as previously described [8] . Dewaxed sections were preincubated for 20 min in buffer containing 50 mM sodium acetate and 40 mM sodium tartrate at pH 5 . 0 . Sections were then incubated for 15 min at room temperature in the same buffer containing 2 . 5 mg/ml naphthol AS-MX phosphate ( Sigma-Aldrich , St . Louis , MO ) in dimethylformamide as substrate and 0 . 5 mg/ml fast garnet GBC ( Sigma-Aldrich ) as a color indicator for the reaction product . After washing with distilled water , the sections were counterstained with methyl green and mounted in Kaiser's glycerol . Osteocalcin , parathyroid hormone-related protein ( PTHrP ) and proliferating cell nuclear antigen ( PCNA ) were determined by immunohistochemistry using the avidin-biotin-peroxidase complex technique with an affinity-purified , goat anti-mouse osteocalcin antibody ( Biomedical Technologies , Inc . , Stoughton MA , USA ) , a mouse anti-PCNA monoclonal antibody ( Medicorp Inc . , Montreal , Canada ) and a rabbit anti-serum against PTHrP[1]–[34] , as described previously [8] , [33] , [47] . Briefly , the primary antibodies were applied to dewaxed paraffin sections overnight at room temperature . As a negative control , pre-immune serum was substituted for the primary antibody . After washing with high salt buffer ( 50 mM Tris–HCl , 2 . 5% NaCl , 0 . 05% Tween 20 , pH 7 . 6 ) for 10 min at room temperature followed by two 10 min washes with TBS , the sections were incubated with secondary antibody ( biotinylated goat anti-mouse IgG , Sigma ) , washed as before and incubated with the Vectastain Elite ABC kit ( Vector Laboratories , Inc . Ontario , Canada ) for 45 min . After washing as before , brown pigmentation to demarcate regions of immunostaining was produced by a 10–15 min treatment with the DAB kit ( Vector Laboratories , Inc . ) . After washing with distilled water , the sections were counterstained with hematoxylin . Sections stained histochemically or immunohistochemically were photographed with a digital camera . Images of micrographs from single sections were digitally recorded using a rectangular template , and recordings were processed and analyzed using Northern Eclipse image analysis software as described [22] , [47] , [49] . Parathyroid gland size was quantitated in H & E stained sections using Northern Eclipse image analysis software and were presented as parathyroid areas ( µm2 ) . RNA was isolated from mouse bone tissues , using Trizol reagent ( Invitrogen ) according to the manufacturer's protocol . Reverse transcription reactions were performed using the SuperScript First-Strand Synthesis System ( Invitrogen ) as previously described [22] , [33] . To determine the number of cDNA molecules in the reverse-transcribed samples , real-time PCR was performed using a LightCycler system ( Roche Molecular Biochemicals , Indianapolis , IN , USA ) . PCR was performed using 2 µl LightCycler DNA Master SYBR Green I ( Roche ) , 12 . 5 µl of reaction mixture , 2 µl of each 5′ and 3′ primer , 2 µl samples and then H2O was added to a final volume of 25 µl . Samples were denatured at 95°C for 10 sec , with a temperature transition rate of 20°C per sec . Amplification and fluorescence determination were carried out in four steps: denaturation at 95°C for 1 sec , with a temperature transition rate of 20°C /sec; annealing for 5 sec , with a temperature transition rate of 8°C /sec; extension at 72°C for 20 sec , with a temperature transition rate of 4°C /sec; and detection of SYBR Green fluorescence , which reflects the amount of double-stranded DNA , at 86°C for 3 sec . The amplification cycle number was 35 . To discriminate specific from nonspecific cDNA products , a melting curve was obtained at the end of each run . Products were denatured at 95°C for 3 sec , and the temperature was then decreased to 58°C for 15 sec and raised slowly from 58 to 95°C using a temperature transition rate of 0 . 1°C /sec . To determine the number of copies of the targeted DNA in the samples , purified PCR fragments of known concentrations were serially diluted and served as external standards that were measured in each experiment . Data were normalized with GAPDH levels in the samples . The primer sequences used for the real-time PCR were the same as those used for routine PCR . Data from image analysis are presented as mean ± SEM . Statistical comparisons were made using a two-way ANOVA , with P<0 . 05 being considered significant . | Mice with homozygous deletion of the calcium-sensing receptor ( CaR ) mimic the syndrome of neonatal severe hyperparathyroidism ( NSHPT ) in humans with very high circulating parathyroid hormone ( PTH ) and severe life-threatening hypercalcemia . To determine effects of CaR deficiency on skeletal development and interactions between CaR and 1 , 25 ( OH ) 2D3 or PTH on calcium and skeletal homeostasis , we compared the skeletal phenotypes of homozygous CaR–deficient mice to those of double homozygous CaR– and 1 , 25 ( OH ) 2D3–deficient mice or those of double homozygous CaR– and PTH–deficient mice . CaR–deficient mice had hypercalcemia , hypophosphatemia , hyperparathyroidism , severe skeletal growth retardation , and abnormalities; and most died within 2 weeks of age . Deletion of 1 , 25 ( OH ) 2D3 in CaR–deficient mice resulted in a longer lifespan , normocalcemia , lower serum phosphorus , greater elevation in PTH , and slight improvement in skeletal growth . Deletion of PTH in CaR–deficient mice resulted in rescue of early lethality , normocalcemia , increased serum phosphorus , and normalization in skeletal growth . Our results indicate that reductions in hypercalcemia reduce the early lethality of CaR–deficient mice and that deletion of PTH in patients with NSHPT may normalize skeletal growth and development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"nutrition",
"endocrinology",
"vitamins",
"diabetes",
"and",
"endocrinology",
"endocrine",
"physiology"
] | 2011 | The Abnormal Phenotypes of Cartilage and Bone in Calcium-Sensing Receptor Deficient Mice Are Dependent on the Actions of Calcium, Phosphorus, and PTH |
Because cohesion prevents sister-chromatid separation and spindle elongation , cohesion dissolution may trigger these two events simultaneously . However , the relatively normal spindle elongation kinetics in yeast cohesin mutants indicates an additional mechanism for the temporal control of spindle elongation . Here we show evidence indicating that S-phase CDK ( cyclin dependent kinase ) negatively regulates spindle elongation . In contrast , mitotic CDK promotes spindle elongation by activating Cdc14 phosphatase , which reverses the protein phosphorylation imposed by S-phase CDK . Our data suggest that S-phase CDK negatively regulates spindle elongation partly through its phosphorylation of a spindle pole body ( SPB ) protein Spc110 . We also show that hyperactive S-phase CDK compromises the microtubule localization of Stu2 , a processive microtubule polymerase essential for spindle elongation . Strikingly , we found that hyperactive mitotic CDK induces uncoupled spindle elongation and sister-chromatid separation in securin mutants ( pds1Δ ) , and we speculate that asynchronous chromosome segregation in pds1Δ cells contributes to this phenotype . Therefore , the tight temporal control of spindle elongation and cohesin cleavage assure orchestrated chromosome separation and spindle elongation .
In eukaryotic cells , spindle elongation and sister-chromatid separation are two critical mitotic events , and the coordination of these two events is essential for the fidelity of chromosome segregation . During mitosis , the cleavage of sister chromatid cohesion allows the sister chromatids to move toward the respective spindle pole , then the spindle elongates further to pull sister chromatids into two daughter cells [1] . Because sister-chromatid cohesion prevents sister chromatid separation and spindle elongation , these two events could be coupled by the cleavage of cohesin . Before anaphase entry , cohesin cleavage is prohibited by the presence of securin ( Pds1 ) , which binds to and inhibits separase ( Esp1 ) that cleaves cohesin Scc1/Mcd1 [2] , [3] . The degradation of Pds1 before anaphase entry alleviates the inhibition of Esp1 , resulting in the robust cohesin cleavage and simultaneous sister-chromatid separation [2] , [4] , [5] . However , yeast cells lacking either Pds1 or cohesin do not show premature spindle elongation , indicating that a cohesion-independent mechanism controls the timing of spindle elongation [6] . CDKs are the key driving force for cell cycle progression . In budding yeast , S-phase cyclins Clb5 and Clb6 appear during S-phase , which is consistent with their function in DNA synthesis . Compared to mitotic CDK Clb2-Cdk1 , Clb5-Cdk1 shows stronger kinase activity toward a subset of CDK substrates [7] . In addition to proteins involved in DNA synthesis , other Clb5-specific substrates , such as Sli15 , Fin1 , Ase1 , and Spc110 , associate with the SPB or microtubules , indicating that S-phase CDK may also regulate spindle dynamics [8]–[10] . The tightly regulated activity of CDK and phosphatase enables unique temporal phosphorylation kinetics of each CDK substrate during the cell cycle . The periodical expression of cyclins controls the activity and substrate specificity of the CDK , while a conserved protein phosphatase Cdc14 reverses the phosphorylation of these CDK substrates [11] . In budding yeast , the regulation of Cdc14 activity is achieved through its subcellular localization . Before anaphase entry , Cdc14 is sequestered within the nucleolus by binding to a nucleolar protein Net1 [12] , [13] . The dephosphorylation of Net1 by PP2ACdc55 ensures a strong Net1-Cdc14 interaction , while during early anaphase the phosphorylation of Net1 by mitotic CDK , Clb2-Cdk1 , triggers the release of Cdc14 from the nucleolus [14]–[17] . The degradation of Pds1 frees separase Esp1 , which not only cleaves cohesin rings but also down-regulates PP2ACdc55 with the assistance of other FEAR components [15] . The increased ratio of Clb2-Cdk1/PP2ACdc55 leads to Net1 phosphorylation and the subsequent Cdc14 release . Because S-phase cyclin Clb5 is also degraded before anaphase entry [18] , the combination of increased Cdc14 and decreased Clb5-Cdk1 activity during early anaphase results in the dephosphorylation of Clb5-specific substrates [19] , [20] . Therefore , mitotic CDK may promote mitotic progression by reversing the phosphorylation imposed by S-phase CDK . Recent work in the Amon lab has shown that the loss of function of both mitotic cyclins Clb1 and Clb2 in clb1Δ clb2-IV mutant cells prevents spindle elongation , indicating the essential role of mitotic CDK in this process . Moreover , neither the loss of sister-chromatid cohesion nor deletion of securin Pds1 is able to rescue the spindle elongation defect in clb1Δ clb2-IV mutant cells , further supporting the direct role of mitotic cyclins in spindle elongation , but the CDK substrates involved in this process remain unknown [21] . If mitotic CDK promote spindle elongation , cells overexpressing these cyclins are expected to show premature spindle elongation , but these cells exhibit relatively normal spindle elongation , although defect in mitotic exit was noticed [22] , [23] . The presence of the CDK inhibitory kinase Swe1 may prevent hyper-activation of mitotic CDK when CLB2 is overexpressed , as Swe1 specifically inhibits mitotic CDK [24]–[27] . Here we show evidence indicating that overexpression of mitotic cyclin CLB2 induces premature spindle elongation in swe1Δ mutant cells . We further found that FEAR mutants suppress this premature spindle elongation , suggesting that Clb2 induces spindle elongation through FEAR that facilitates Cdc14 release during early anaphase . In contrast to mitotic cyclins , we found that S phase cyclin Clb5 plays a negative role in spindle elongation . Therefore , mitotic CDK likely activates the FEAR pathway to alleviate the inhibitory effect of S-phase CDK on spindle elongation . Our data further suggest that the phosphorylation of a SPB protein Spc110 by S-phase CDK contributes at least partially to the inhibition of spindle elongation . Moreover , high levels of S-phase CDK prevent microtubule localization of Stu2 , a microtubule-plus-end tracking protein essential for spindle elongation . Strikingly , overexpression of CLB2 in securin mutants pds1Δ leads to uncoupled sister-chromatid separation and spindle elongation . Given the established role of securin Pds1 in the synchronous chromosome segregation [5] , our data support the possibility that this securin-dependent synchrony and the temporal control of spindle elongation by the balance of mitotic versus S-phase CDK ensure the sequential order of sister-chromatid separation and spindle elongation , which is critical for faithful chromosome segregation .
Clb2 is the major mitotic cyclin in budding yeast , but its overexpression from a galactose-inducible promoter does not cause obvious premature mitosis in wild-type ( WT ) cells . Because Swe1 kinase phosphorylates and inhibits mitotic CDK , it is possible that the presence of Swe1 prevents the hyper-activation of mitotic CDK after CLB2 overexpression . Therefore , we overexpressed CLB2 from a galactose-inducible promoter in swe1Δ mutant cells and examined the cell growth . The Western blotting result confirmed the high level expression of Clb2 protein after galactose induction ( Figure S1A ) . Compared to the control cells , swe1Δ mutants with a PGALCLB2 plasmid showed obvious growth defect on galactose plates . Overexpression of Clb1 , which is closely related to Clb2 , also caused sick growth phenotype in swe1Δ cells , but overexpression of Clb3 , Clb4 , or S phase cyclin Clb5 , Clb6 was not toxic ( Figure 1A and S1B ) . Therefore overexpression of mitotic cyclins Clb1 and Clb2 is toxic to swe1Δ cells . To confirm that mitotic CDK is hyperactive in swe1Δ cells after CLB2 overexpression , we compared the phosphorylation kinetics of Pol12 in synchronized WT and swe1Δ cells overexpressing CLB2 , as Pol12 is a known substrate of mitotic CDK required for DNA replication [24] . Our results showed that CLB2 overexpression induces premature Pol12 phosphorylation in both WT and swe1Δ cells based on the band-shift , and the phosphorylation became more significant in swe1Δ cells as indicated by the increased slow migrating band ( Figure S1C ) . Therefore , the absence of Swe1 indeed causes hyper-activation of mitotic CDK after CLB2 overexpression . To understand the cause of this sick growth phenotype , G1-arrested WT and swe1Δ cells with a control vector or a PGALCLB2 plasmid were released into galactose medium to induce CLB2 overexpression and we compared the cell cycle progression in these cells . Both WT and swe1Δ cells showed almost identical budding index and DNA synthesis kinetics either with or without CLB2 overexpression ( Figure S2A and S2B ) . These cells also exhibited similar cell cycle regulated fluctuation of Pds1 protein levels ( Figure S2C ) . However , we noticed premature spindle elongation in some small-budded and unbudded swe1Δ cells overexpressing CLB2 , but this phenotype is much less significant in WT cells ( Figure 1B ) . After G1 release into galactose medium for 120 min , 14% of WT cells with a PGALCLB2 plasmid had elongated spindles ( >3 µM ) , while 36% of swe1Δ cells showed elongated spindles . Interestingly , about 8% swe1Δ cells became binucleate after G1 release for 160 min , i . e . two nuclei were observed in a single cell body ( the arrow in Figure 1B ) . Among them , half were small-budded , while the others were unbudded . Previous data indicate that overexpression of a single copy of a Clb2 destruction box mutant prevents bud formation and results in binucleate cells [28] . Indeed , we found that some cells overexpressing CLB2 remained unbudded after G1 release ( Figure S2A ) . To further define the role of the inhibition of budding and premature spindle elongation in the formation of binucleate cells , we performed live-cell imaging to examine the dynamics of spindle elongation . G1-arrested swe1Δ cells with a control vector or a PGALCLB2 plasmid were released into galactose medium . The spindle elongation in swe1Δ ( PGALCLB2 ) cells initiated ∼20 min earlier compared to the cells with a control vector . Interestingly , we observed premature spindle elongation in both small-budded and unbudded swe1Δ cells ( Figure 1C ) , indicating that both budding inhibition and premature spindle elongation may contribute to the formation of binucleate cells . Because we also observed the binucleate phenotype in both unbudded and small budded cells after Clb2 overproduction ( Figure 1B ) , we reason that the inhibition of budding is not essential for the formation of binucleate cells . If a cell elongates spindle when cohesion is still present , some sister chromatids may remain together after spindle elongation . However , all the swe1Δ cells overexpressing CLB2 with an elongated spindle showed separated sister chromatids ( Figure S3A ) , indicating that spindle elongation did not occur prior to cohesion dissolution . Alternatively , hyperactive mitotic CDK may promote cohesin cleavage . Thus , we examined Scc1 proteins in WT and swe1Δ cells with and without CLB2 overexpression , but all these cells exhibited similar Scc1 cleavage kinetics based on the appearance of the short Scc1 fragments ( Figure S3B ) , arguing against the possibility that Clb2 induces spindle elongation through cohesin cleavage . We speculate that both hyperactive mitotic CDK and cohesion dissolution are essential for spindle elongation . If that is the case , overexpression of CLB2 may cause more dramatic premature spindle elongation in cohesin mutant cells . We first found that scc1-73 mutant cells with PGALCLB2 plasmids grew more slowly on galactose plates at 25°C compared to control cells ( Figure S4A ) . Moreover , after G1 release , CLB2 overexpression caused premature spindle elongation in scc1-73 cells ( Figure S4B ) , and this phenotype became more pronounced in swe1Δ scc1-73 cells ( Figure S4C ) . Therefore , hyperactive mitotic CDK cause more dramatic premature spindle elongation in cells with compromised sister chromatid cohesion . One of the substrates of mitotic CDK is the nucleolar Cdc14-binding protein Net1 , whose phosphorylation triggers the dissociation of Cdc14 from Net1 and the release of Cdc14 from the nucleolus . It is possible that hyperactive mitotic CDK stimulates spindle elongation by activating FEAR . Because the replacement of 6 CDK phosphorylation sites in Net1 with alanine generates net1-6Cdk mutant , which prevents FEAR activation [14] , we first compared the growth of swe1Δ and swe1Δ net1-6Cdk cells after CLB2 overexpression . The swe1Δ net1-6Cdk cells grew much better than the single mutant cells after CLB2 overexpression . Another FEAR mutant spo12Δ showed an even stronger suppression of the sick growth phenotype of swe1Δ ( Figure 2A ) . The nuclear morphology was also compared in swe1Δ , swe1Δ spo12Δ and swe1Δ net1-6Cdk mutant cells overexpressing CLB2 , both spo12Δ and net1-6Cdk mutants suppressed the formation of binucleate cells ( Figure 2B ) , suggesting that the activation of FEAR pathway contributes to the growth defect in swe1Δ cells overexpressing CLB2 . To directly determine if the toxicity of CLB2 overexpression to swe1Δ cells is due to hyperactive Cdc14 , we examined the growth of swe1Δ cdc14-1 cells overexpressing CLB2 . We found that cdc14-1 swe1Δ cells with PGALCLB2 plasmids grew better than swe1Δ cells on galactose plates at 25°C ( Figure 2A ) . In addition , the cdc14-1 mutant partially suppressed binucleate phenotype of swe1Δ cells ( Figure 2C ) . We reason that the incomplete suppression is due to the presence of partial functional Cdc14 in cdc14-1 mutant . Therefore , CLB2 overexpression likely induces premature spindle elongation by activating Cdc14 . Unlike the FEAR pathway , the mitotic exit network ( MEN ) induces Cdc14 release in late anaphase [22] , [29] . To further test if Clb2-Cdk1 promotes spindle elongation by activating FEAR or MEN , we examined Clb2-induced premature spindle elongation in cdc15-2 swe1Δ cells . The abolishment of MEN function in cdc15-2 mutant did not block Clb2-induced premature spindle elongation . However , the introduction of either net1-6Cdk or spo12Δ mutation in swe1Δ cdc15-2 cells abolished premature spindle elongation completely ( Figure 2D and Figure S5 ) , indicating that Clb2-Cdk1 induces this phenotype through FEAR but not MEN . To directly determine if CLB2 overexpression in swe1Δ cells causes premature FEAR activation , the localization of Cdc14 was examined after G1-arrested swe1Δ cells were released into galactose medium containing microtubule poison nocodazole that arrests cells in metaphase . Strikingly , 75% of swe1Δ cells with a PGALCLB2 plasmid showed released Cdc14 after incubation in galactose medium for 4 hrs , while only 23% of WT cells showed this phenotype . In contrast , swe1Δ cells with a control vector did not show any Cdc14 release . net1-6Cdk mutant suppressed the premature Cdc14 release in swe1Δ cells ( Figure 2E ) . All these data indicate that excess mitotic CDK likely trigger premature spindle elongation by activating Cdc14 through the FEAR pathway . However , we cannot exclude the possibility that Clb2-Cdk1 also phosphorylates other substrates to promote spindle elongation . Overexpression of CLB5 slows cell growth , suggesting that hyperactive Clb5-Cdk1 may have a negative effect on the cell cycle ( Figure 1A ) . To test if hyperactive S-phase CDK inhibits spindle elongation , we examined the spindle structure in WT cells overexpressing CLB5 . After incubation in galactose medium for 4 hrs , more yeast cells with a PGALCLB5 plasmid arrested with a large bud and a very short spindle structure ( Figure 3A ) . Because these spindles are very short , one explanation is that the failure of SPB separation contributes to the spindle elongation defect . Therefore , we examined the spindle elongation kinetics in cells overexpressing CLB5 after release from hydroxyurea ( HU ) arrest . HU blocks DNA synthesis and HU-arrested cells have a short spindle with separated SPBs [30] . After HU wash off , CLB5 overexpression also caused a clear spindle elongation delay as indicated by the accumulation of cells with a bar-like short spindle structure ( Figure 3B ) . This result suggests that the short spindle observed in cells with hyperactive S-phase CDK is not due to SPB separation defect . As we have shown that CLB5 overexpression blocks nuclear division in FEAR mutants , such as slk19Δ and spo12Δ [19] , we further examined the spindle elongation kinetics in spo12Δ mutants with and without CLB5 overexpression after HU release . Obviously , spindle elongation was largely blocked by CLB5 overexpression in spo12Δ mutant cells ( Figure 3B ) . The spindle elongation defect could be due to the activation of the DNA damage or the spindle checkpoint that prevents anaphase onset . Because the activation of these checkpoints depends on the stabilization of securin Pds1 [31] , [32] , deletion of PDS1 will abolish these checkpoints . We therefore compared the spindle elongation kinetics in WT , pds1Δ , spo12Δ , and spo12Δ pds1Δ mutant cells when CLB5 is overexpressed . Like WT and spo12Δ mutants , pds1Δ and spo12Δ pds1Δ mutants also exhibited accumulation of large-budded cells with a short spindle structure after CLB5 overexpression ( Figure 3C ) , indicating that hyperactive S-phase CDK prevents spindle elongation in a checkpoint-independent manner . Another possibility is that high levels of S-phase CDK block the cleavage of sister chromatid cohesin to prevent spindle elongation . Strikingly , cells overexpressing CLB5 showed almost identical kinetics for cohesin cleavage compared to control cells ( Figure S6A ) . Moreover , overexpression of CLB5 also caused delayed spindle elongation in cohesin mutants incubated at the non-permissive temperature ( Figure S6B ) . Therefore , we conclude that S-phase CDK negatively regulates spindle elongation , and this function is unlikely due to the inability of cohesion resolution . If S-phase CDK plays a negative role in spindle elongation , we expect that the loss of S-phase cyclins will cause premature spindle elongation . However , both clb5Δ and clb5Δ clb6Δ mutants showed normal spindle elongation kinetics . It is possible that other redundant mechanism , such as the presence of sister-chromatid cohesion , prevents premature spindle elongation in these mutant cells . To test this possibility , we need to examine the spindle elongation kinetics in clb5Δ clb6Δ mutants in the absence of cohesion . Because the spindle is unstable in cohesin mutant cells and deletion of the spindle checkpoint , such as MAD2 , suppresses the spindle instability [33] , we determined the spindle elongation kinetics in scc1-73 mad2Δ and clb5Δ clb6Δ scc1-73 mad2Δ mutants . Consistent with previous data , scc1-73 mad2Δ cells elongated spindle with kinetics similar to WT cells , but clb5Δ clb6Δ scc1-73 mad2Δ mutant cells showed premature spindle elongation ( Figure S7A ) . Therefore , the absence of S-phase cyclins leads to premature spindle elongation in the absence of cohesion . Our data indicate that S-phase and mitotic CDK may play opposing roles in spindle elongation . If that is true , we expect CLB2 overexpression to cause premature spindle elongation in the absence of S-phase cyclins . Therefore , we examined spindle elongation in clb5Δ clb6Δ mutant cells with a PGALCLB2 plasmid and control cells . As expected , clb5Δ clb6Δ cells showed premature spindle elongation when CLB2 is overexpressed and these cells grew slowly on galactose medium ( Figure S7B ) . Together , these data support the conclusion that S-phase CDK plays a negative role in spindle elongation , while mitotic CDK plays a positive role in this process . Functional FEAR is required for Clb2-induced premature spindle elongation , and the FEAR promotes Cdc14 release to dephosphorylate Clb5 substrates , such as Ase1 and Fin1 [7] , [34] , but we found that ase1Δ or fin1Δ mutant did not suppress Clb2-induced premature spindle elongation . A previous study suggests that Clb5-Cdk1-induced phosphorylation of Spc110 , one of the SPB proteins , also modulates spindle dynamics [9] . Interestingly , a phospho-mimetic spc11018D91D mutant , in which the CDK phosphorylation sites at Thr18 and Ser91 were mutated to aspartic acid , showed dramatic suppression of the binucleate phenotype in swe1Δ cells overexpressing CLB2 ( Figure 4A ) , indicating that the dephosphorylation of Spc110 might be required for Clb2-induced premature spindle elongation . We further compared the spindle elongation kinetics in WT and spc11018D91D mutant cells and found that the mutant cells did exhibit delayed spindle elongation , although the delay was not pronounced ( Figure S8 ) . Nevertheless , the spc11018D91D mutant failed to rescue the sick growth phenotype of swe1Δ cells with PGALCLB2 on galactose medium , suggesting that phosphorylation of other Clb5 substrates can prevent spindle elongation as well . Alternatively , other unidentified defects induced by Clb2 overexpression may also lead to the sick growth . The phosphorylation of Spc110 by Clb5-Cdk1 produces a band-shift on protein gels detectable by Western blotting [9] . We have demonstrated that the dephosphorylation of some Clb5-Cdk1 substrates depends on functional FEAR [19] . To test if the FEAR pathway is also required for the dephosphorylation of Spc110 , we compared the band-shift of Spc110 protein in WT and mutant cells lacking functional MEN or FEAR . Significant Spc110 phosphorylation was not observed in synchronous WT and MEN mutant cells cdc15-2 , but we noticed more obvious Spc110 phosphorylation in cdc15-2 spo12Δ mutant wherein the function of both MEN and FEAR is abolished , indicating that functional FEAR may be required for Spc110 dephosphorylation . Because mitotic CDK activates the FEAR by phosphorylating Net1 , we also examined Spc110 phosphorylation in clb1Δ clb2-VI temperature sensitive mutants . These mutant cells exhibited more Spc110 phosphorylation , which supports the notion that mitotic CDK promotes Spc110 dephosphorylation ( Figure 4B ) . We previously showed that overexpression of S-phase cyclin CLB5 is toxic to FEAR mutants and delays nuclear separation in the mutant cells [19] . If Clb5-induced phosphorylation of Spc110 contributes to this phenotype , a nonphosphorylatable spc110 mutant will suppress this phenotype . We first found that a nonphosphorylatable spc11018A91A mutant partially restored the growth of FEAR mutants ( spo12Δ ) with a PGALCLB5 plasmid on galactose medium ( Figure 4C ) . The spindle elongation dynamics was also examined in spo12Δ and spo12Δ spc11018A91A mutants overexpressing CLB5 . Clb5 overproduction in spo12Δ mutants significantly delayed spindle elongation . After G1 release for 150 min , 45% of WT cells exhibited elongated spindles , while only 23% of spo12Δ mutant cells did . However , 35% of spo12Δ spc11018A91A mutant cells displayed elongated spindle at 150 min ( Figure 4D ) , indicating that active S-phase CDK prevents spindle elongation at least partially through Spc110 phosphorylation . Consistently , more significant Spc110 phosphorylation was observed in spo12Δ cells overexpressing CLB5 ( Figure 4E ) . In these cells , the kinetics of DNA synthesis is indistinguishable with or without CLB5 overexpression based on the FACS analysis . Collectively , these data reveal the possibility of a negative role of Clb5-dependent Spc110 phosphorylation in spindle elongation . Our data suggest that the phosphorylation of SPB component Spc110 plays a role in the timing control of spindle elongation , and this phosphorylation is regulated by the balance of S-phase and mitotic CDKs . As a SPB component , however , it is likely that the phosphorylation of Spc110 regulates the spindle elongation via other microtubule-associated protein ( s ) . Stu2 is the yeast homologue of the XMAP215 protein that binds to the microtubule plus-end [35] , [36] . This protein is a processive microtubule polymerase essential for spindle elongation [37] , [38] . One possibility is that the CDK activity controls the timing of spindle elongation by regulating the activity of Stu2 . Interestingly , the temperature sensitive mutant stu2-10 dramatically suppressed the toxicity of CLB2 overexpression to swe1Δ mutant cells when incubated at 25°C ( Figure 5A ) , indicating that intact Stu2 function is required for CLB2-induced premature spindle elongation . In contrast , stu2-10 mutant cells were more sensitive to CLB5 overexpression than WT cells ( Figure 5B ) , indicating that Clb5 may negatively regulates Stu2 function . As we have showed that the phosphorylation of Spc110 by Clb5-Cdk1 plays a negative role in spindle elongation ( Figure 4C and 4D ) , we further compared the growth and spindle elongation in stu2-10 and stu2-10 spc11018A91A at 35°C . The results showed that nonphosphorylatable spc110 mutant partially suppressed the temperature sensitivity and the spindle elongation defect of stu2-10 mutant cells ( Figure 5C ) , suggesting that S-phase CDK-dependent Spc110 phosphorylation may down-regulate Stu2 function . To further test if the spindle localization of Stu2 is impaired in cells overexpressing CLB5 , live-cell imaging was performed in cells with and without CLB5 overexpression . Consistent with previous reports , we found that Stu2 localized on both spindle poles and spindle at metaphase and early anaphase [36] . At some cell cycle stages , the localization of Stu2 on the cytoplasmic microtubules was also clearly observed in the control cells ( Figure 5D ) . Although the SPB-localization of Stu2 remained similar , we noticed that overexpression of CLB5 significantly decreased the Stu2 localization on spindle and cytoplasmic microtubules . However , spc11018A91A mutant restored the microtubule-localization of Stu2 in cells overexpressing CLB5 ( Figure 5D ) , which is consistent with the result that spc11018A91A partially suppressed the temperature sensitivity of stu2-10 . Therefore , we speculate that the phosphorylation of Spc110 by S-phase CDK prevents the localization of Stu2 on spindle and cytoplasmic microtubules , presumably at the plus-ends . In contrast , Spc110 dephosphorylation likely facilitates the microtubule localization of Stu2 , which promotes spindle elongation . Our data indicate that the balance of mitotic and S-phase CDKs regulates the timing of spindle elongation . In addition , the presence of sister chromatid cohesion prevents spindle elongation . Although cohesion mutant cells ( scc1-73 ) exhibit premature spindle elongation when CLB2 is overexpressed , most of the cells have a successful mitosis because they are viable after CLB2 overexpression . We suspect that an additional mechanism also plays a role in the coordination of cohesion cleavage and spindle elongation . Securin Pds1 binds to and inhibits separase Esp1 , whose activity is required for the cleavage of cohesion Scc1/Mcd1 and the subsequent anaphase onset [2] . Thus , cell cycle-regulated Pds1 protein levels control the timing of cohesion cleavage and anaphase onset . A previous study showed that deletion of a nonessential gene CDH1 caused lethality in pds1Δ cells and expression of an extra copy of SWE1 suppressed this lethality [39] . Because CDH1 encodes an APC activator required for Clb2 degradation , it is possible that the high Clb2 levels contribute to the synthetic lethality of cdh1 pds1 mutants . Therefore , we examined the growth of pds1Δ cells overexpressing CLB2 and found that overexpression of CLB2 was very toxic to pds1Δ cells ( Figure 6A ) . One possibility is that high levels of mitotic CDK cause chromosome biorientation defects [40] , which require the intact spindle assembly checkpoint for survival . However , overexpression of CLB2 was not toxic to checkpoint mutant mad2Δ ( Figure S9 ) , excluding the possibility that the checkpoint defect in pds1Δ contributes to the lethality . After CLB2 expression for 4 hrs , 68% of pds1Δ cells lost viability ( Figure 6B ) . A FEAR mutant spo12Δ failed to suppress the growth defect of pds1Δ ( PGALCLB2 ) cells on galactose plates , but it partially suppressed the lethality of pds1Δ ( PGALCLB2 ) cells after incubation in liquid galactose media ( Figure 6B ) . Therefore , we conclude that high level of Clb2 proteins causes the lethality in pds1Δ mutant cells . The presence of Pds1 prevents the activation of separase Esp1 and the abrupt Pds1 degradation prior to anaphase onset allows a robust cohesion cleavage , resulting in synchronous dissolution of all chromosome pairs . The absence of Pds1 decreases this synchrony [5] , thus , the loss of this synchrony may cause catastrophic mitosis when CLB2 is overexpressed . To test this possibility , we first examined the spindle elongation and sister centromere separation in pds1Δ cells overexpressing CLB2 . After pds1Δ cells were released from G1-arrest into galactose medium for 3 hrs , about 12% cells exhibited an elongated spindle with a single CEN4-GFP dot , indicating the failure of chromosome IV separation after spindle elongation ( Figure 6C ) . Given that this number is only for one of the 16 chromosomes , the defect in chromosome segregation should be very dramatic and quick viability loss supports this speculation . We found that the mis-segregation of CEN4-GFP was largely suppressed by spo12Δ , which is consistent with the notion that CLB2 promotes spindle elongation through FEAR ( Figure 6C ) . We further used DAPI staining to examine the chromosome segregation in pds1Δ cells overexpressing CLB2 . Strikingly , most of the cells failed to show two clear DNA masses after spindle elongation . Instead , they exhibited lagged DNA along the elongated spindle ( Figure 6D ) , indicating a remarkable chromosome segregation defect . The observation of cells with an elongated spindle and a single CEN4-GFP dot supports this speculation . We further examined the segregation of the telomere of chromosome V ( TEL5-GFP ) in pds1Δ cells overexpressing CLB2 . After G1 release into galactose for 180 min , 23% of pds1Δ cells exhibited a single TEL5-GFP dot with two kinetochore clusters ( Nuf2-mCherry ) separated to two daughter cells , indicating the failure of telomere separation for chromosome V ( Figure 6E ) . spo12Δ also showed partial suppression of TEL5-GFP mis-segregation ( Figure 6E ) . Interestingly , the percentage of pds1Δ cells with a single TEL5-GFP dot after CLB2 overexpression is obviously higher than that with unseparated CEN4-GFP ( 23% vs . 11% ) . Our explanation is that both unseparated and partially separated chromosomes contribute to the telomere separation defect . The results support the possibility that overexpression of CLB2 in pds1Δ cells induces spindle elongation when cohesin still holds a few chromosomes or some parts of a chromosome . Nevertheless the small amount of cohesin is unable to restrain the premature spindle elongation induced by CLB2 overexpression , thereby resulting in the failure for the segregation of entire or part of a chromosome . Premature spindle elongation in pds1Δ mutant cells , therefore , induces uncoupled chromosome segregation and spindle elongation , which leads to significant chromosome mis-segregation .
The key to a successful cell division is the coordination of various cell cycle events . For efficient chromosome segregation , spindle elongation should follow the dissolution of sister-chromatid cohesion in an orderly fashion . The molecular mechanism that ensures this sequential order remains unclear . The absence of premature spindle elongation in cells lacking cohesion indicates a cohesion-independent mechanism that controls the timing of spindle elongation . Here we show that S-phase CDK negatively regulates spindle elongation , while mitotic CDK actives the FEAR pathway to trigger Cdc14 release , which reverses S-phase CDK-dependent protein phosphorylation and simulates spindle elongation . Therefore , the balance of mitotic vs . S-phase CDK activity is critical for the timing control of spindle elongation . We also show that S-phase CDK prevents spindle elongation in part by phosphorylating a SPB component Spc110 , while dephosphorylation of Spc110 by Cdc14 likely facilitates the localization of Stu2 , a plus-end tracking protein , to spindle microtubules , which may directly promotes spindle elongation by enhance microtubule polymerization [37] . Furthermore , hyperactive mitotic CDK in pds1Δ cells , where the synchrony of chromosome segregation is compromised , leads to uncoupled sister-chromatid separation and spindle elongation , resulting in chromosome mis-segregation and cell death . A model illustrating and integrating this cell cycle regulatory network is shown in Figure 7 . These findings reveal at least two mechanisms that prevent premature spindle elongation . Firstly , before anaphase onset , S-phase CDK phosphorylates several microtubule-associated proteins to prevent spindle elongation . Therefore , in addition to promoting DNA synthesis , S-phase CDK also negatively regulates mitosis to ensure the correct order of S and M phase spindle function [41] . Secondly , sister-chromatid cohesion restrains spindle elongation as well . To allow spindle elongation , both the loss of sister-chromatid cohesion and the reversion of protein phosphorylation imposed by S-phase CDK have to be achieved . Before anaphase onset , the destruction of Pds1 frees separase Esp1 that cleaves cohesin rings . Nevertheless , the reversion of S-phase CDK-dependent phosphorylation requires the destruction of S-phase cyclin Clb5 as well as the activation of phosphatase Cdc14 . Clb5 is degraded before anaphase onset along with Pds1 through APCCdc20 [18] . The release of Esp1 from the inhibition by Pds1 not only triggers cohesin cleavage but also activates FEAR to release Cdc14 [15] . Therefore , this mechanism ensures that cohesin cleavage and spindle elongation are coordinated temporally ( Figure 7 ) . The tightly regulation of the protein levels of securin Pds1 during cell cycle not only avoids the dissolution of sister-chromatid cohesion prior to anaphase entry [42] , but also ensures that all chromosome pairs disjoin almost simultaneously [5] . Loss of this switch-like mechanism in pds1Δ mutant cells results in asynchronous cohesion dissolution . Therefore , in pds1Δ mutant cells , it is possible that only a few sister chromatids are linked by cohesin to restrain spindle elongation while cells have initiated spindle elongation . Surprisingly , most pds1Δ cells are able to segregate chromosomes successfully . The temporal control mechanism described above may prevent spindle elongation when cohesion is partially dissolved in pds1Δ mutants , but deregulation of this temporal control mechanism could lead to catastrophic mitosis . Indeed , we showed that overexpression of CLB2 in pds1Δ mutants results in the failure of the separation of some sister chromatids after spindle elongation . In addition to high levels of mitotic cyclins , constitutive activation of FEAR by deletion of CDC55 or the deletion of S-phase cyclins has also been shown to be lethal to pds1Δ mutants [39] , [43] , [44] , and the induction of premature spindle elongation is likely the cause of the lethality . Because only some sister-chromatids are linked by cohesin rings when spindles elongate in these cells , the pulling force may break the kinetochore-microtubule interaction , resulting in chromosome mis-segregation . It is also possible that cohesin only exists in part of a chromosome , such as the arm or telomere regions while the spindle is elongating . Under this situation , sister centromeres segregate successfully , but the telomeres of this chromosome still stay together . Previous work validates the role of mitotic CDK in spindle elongation , as cells lacking both Clb1 and Clb2 fail to elongate the spindle . Because overexpression of Cdc14 phosphatase cannot suppress the spindle elongation defect in clb1Δ clb2-IV mutant , it was speculated that mitotic CDK promotes spindle elongation in FEAR-independent manner [21] . Our observation that FEAR mutants completely suppress Clb2-induced premature spindle elongation and toxicity to swe1Δ cells strongly supports the conclusion that mitotic CDK promotes spindle elongation through the FEAR pathway . However , we cannot exclude the possibility that additional mitotic CDK targets might also be involved in spindle elongation . An interesting question is whether cohesin cleavage and spindle elongation occur at the same time . We found that spindle elongation happens earlier in swe1Δ mutant cells overexpressing CLB2 , but the cells with elongated spindle always show separated sister chromatids . Moreover , our result indicates that CLB2 overexpression does not leads to premature cohesin cleavage . These results suggest that spindle elongation may occur later than cohesin cleavage , but hyperactive CDK eliminates this lag . Because FEAR mutants block Clb2-induced premature spindle elongation in swe1Δ mutants , we speculate that the activation of FEAR occurs later than cohesin cleavage , which contributes to the time difference . Therefore , an important function of the FEAR pathway is likely to ensure the sequential order of cohesin cleavage and spindle elongation . Although the hyperactive FEAR in cdc55Δ mutant cells did not result in dramatic mitotic defect , cdc55Δ bub2Δ double mutant cells exhibit catastrophic mitosis , wherein both FEAR and MEN are hyperactive [16] . Several microtubule-associated proteins , such as Ase1 , Fin1 , and Ask1 , are phosphorylated more efficiently by S-phase CDK and their phosphorylation regulates spindle dynamics [8] , [34] , [45] . Here we show that S-phase CDK-mediated phosphorylation of Spc110 also plays an important role in spindle elongation . We found that the phospho-mimetic spc11018D91D mutant partially abolishes Clb2-induced premature spindle elongation . In contrast , the nonphosphorylatable spc11018A91A mutant suppresses the Clb5-induced delay in spindle elongation . Although spc11018D91D mutant cells show a noticeable delay in spindle elongation , the delay is not significant . We reason that the phosphorylation of a group of proteins by Clb5-Cdk1 prevents spindle elongation , but the phospho-mimetic mutant for each single protein may not be sufficient to block spindle elongation completely . In contrast to Ase1 and Fin1 , which directly bind to microtubules and regulate microtubule dynamics , Spc110 is a SPB component , so it is likely that Spc110 phosphorylation regulates spindle elongation in an indirect way . We show that overexpression of CLB5 clearly impairs the localization of Stu2 on spindle and cytoplasmic microtubules . The dephosphorylation of Spc110 and/or other Clb5 substrates likely promotes the localization of Stu2 to the plus-end of spindle microtubules , where Stu2 triggers the polymerization and spindle elongation . Further studies are needed to understand how S-phase CDK prevents microtubule localization of Stu2 . In summary , our data reveal the molecular mechanism that coordinates chromosome segregation and spindle elongation . Before anaphase , S-phase CDK and sister chromatid cohesion prevent spindle elongation . The establishment of chromosome bipolar attachment triggers the degradation of both securin Pds1 and S-phase cyclin Clb5 . The disappearance of Pds1 activates Esp1 that cleaves cohesin and triggers Cdc14 release through FEAR . Consequently , the loss of sister chromatid cohesion and the reversion of S-phase CDK-dependent protein phosphorylation trigger spindle elongation . This mechanism becomes essential in cells with decreased synchrony of chromosome segregation , as induction of premature spindle elongation in these cells results in unseparated or partially separated chromosomes after spindle elongation . Like budding yeast Clb5 , the S-phase cyclin ( Cyclin A ) in mammalian cells also exhibits substrate specificity [46] . Moreover , the proteins involved in this regulation , such as Ase1 , Spc110 , and Stu2 , are well conserved in higher eukaryotic cells . For example , XMAP215 , the Stu2 homologue in mammalian cells , is required for microtubule polymerization and spindle elongation [47] . Therefore , the mechanism of the temporal control of spindle elongation described in budding yeast could be conserved . We represent data showing that the temporal control of spindle elongation is critical for accurate chromosome segregation , suggesting that defects in this network may contribute to aneuploidy that is associated with cancer progression .
The yeast strains used in this study are listed in Table S1 . All strains are isogenic to Y300 , a W303 derivative . Yeast cells were grown in YPD ( Yeast extract , Peptone , Dextrose ) or indicated synthetic medium . To arrest cells in G1 phase , 5 µg/ml α-factor was added into cell cultures . After 2 hr incubation , the G1-arrested cells were washed twice with water and then released into fresh medium to start the cell cycle . Nocodazole was used at 20 µg/ml in 1% DMSO . To induce cyclin overexpression , galactose was added to the medium to a final concentration of 2% . Cells with GFP , mCherry or mApple-tagged proteins were fixed with 3 . 7% formaldehyde for 5 min at room temperature , and then washed twice with 1× PBS buffer and resuspended in PBS buffer for fluorescence microscopy ( Carl Zeiss MicroImaging , Inc . ) . The spindle morphology was monitored by using TUB1-GFP , TUB1-mApple , or TUB1-mCherry strains and we count the spindles longer than 3 µm as elongated spindles . For DAPI staining , cells were fixed with 3 . 7% formaldehyde for 5 min at room temperature , and then resuspended in 100% MEOH at −20°C for 30 min . The cells were incubated in DAPI solution ( final concentration of 2 . 5 µg/ml ) for 1 min at room temperature . For each experiment , we repeated 3 times and at least 100 cells were counted for each sample . Live-cell microscopy was carried out with a Nikon Eclipse Ti imaging system ( Andor ) . We used a glass depression slide to prepare an agarose pad filled with synthetic complete medium with the addition of galactose . All live-cell images were acquired at 25°C with an ×100 objective lens . Twelve Z-sections were collected at each time point , and each optical section was set at 0 . 5 µm thick . The time-lapse interval was set at 5 min . Maximum projection from applicable time points were created using Andor IQ2 software . G1 phase-arrested cells in raffinose medium were released into galactose medium . Samples were taken at various time points and fixed in 70% EtOH overnight at 4°C . Cells were then incubated in Tris pH 7 . 8 buffer with 0 . 2 mg/ml RNase A at 37°C for 4 hrs and stained with 30 µg/ml propidium iodide at 4°C overnight . FACS analysis was performed using FACSCanto equipped with the FACSDiva software . Cell pellets from 1 . 5 ml of cell culture were resuspended in 200 µl 0 . 1 N NaOH and incubated at room temperature for 5 min . After centrifugation , the cells were resuspended in equal volume ( 30 µl ) of ddH2O and SDS protein-loading buffer . The samples were then boiled for 5 min and resolved with 8% SDS-polyacrylamide gel . Proteins were detected with ECL ( Perkin Elmer LAS , Inc . ) after probing with anti-Myc or anti-HA antibodies ( Covance Research Products , Inc . ) and HRP-conjugated secondary antibody ( Jackson ImmunoResearch , Inc . ) . | Before anaphase onset , the presence of securin Pds1 prevents the cleavage of cohesin , a protein complex that holds sister chromatids together , but the degradation of Pds1 leads to robust cohesion dissolution , resulting in synchronous separation of all chromosomes . Since sister chromatids are attached to spindle microtubules before anaphase onset , spindle elongation segregates sister chromatids to each daughter cell after cohesion dissolution . A fundamental question is how spindle elongation is temporally regulated in order to coordinate with sister chromatid separation . Cyclin-dependent kinases ( CDKs ) play a key role in the cell cycle , and it is well established that S-phase CDK promotes DNA synthesis . We found that hyperactive S-phase CDK delays spindle elongation in budding yeast , while hyperactive mitotic CDK leads to premature spindle elongation . We further provide evidence indicating that mitotic CDK promotes spindle elongation by activating a protein phosphatase that antagonizes S-phase CDK , thus the balance of S-phase and mitotic CDK controls the timing of spindle elongation . Interestingly , hyperactive mitotic CDK induces uncoupled spindle elongation and sister chromatid separation in the absence of securin , suggesting that the temporal control of spindle elongation and cohesin cleavage is essential for faithful chromosome segregation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"model",
"organisms",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2013 | Coordination of Chromatid Separation and Spindle Elongation by Antagonistic Activities of Mitotic and S-Phase CDKs |
The thioredoxin and glutaredoxin pathways are responsible of recycling several enzymes which undergo intramolecular disulfide bond formation as part of their catalytic cycles such as the peroxide scavengers peroxiredoxins or the enzyme ribonucleotide reductase ( RNR ) . RNR , the rate-limiting enzyme of deoxyribonucleotide synthesis , is an essential enzyme relying on these electron flow cascades for recycling . RNR is tightly regulated in a cell cycle-dependent manner at different levels , but little is known about the participation of electron donors in such regulation . Here , we show that cytosolic thioredoxins Trx1 and Trx3 are the primary electron donors for RNR in fission yeast . Unexpectedly , trx1 transcript and Trx1 protein levels are up-regulated in a G1-to-S phase-dependent manner , indicating that the supply of electron donors is also cell cycle-regulated . Indeed , genetic depletion of thioredoxins triggers a DNA replication checkpoint ruled by Rad3 and Cds1 , with the final goal of up-regulating transcription of S phase genes and constitutive RNR synthesis . Regarding the thioredoxin and glutaredoxin cascades , one combination of gene deletions is synthetic lethal in fission yeast: cells lacking both thioredoxin reductase and cytosolic dithiol glutaredoxin . We have isolated a suppressor of this lethal phenotype: a mutation at the Tpx1-coding gene , leading to a frame shift and a loss-of-function of Tpx1 , the main client of electron donors . We propose that in a mutant strain compromised in reducing equivalents , the absence of an abundant and competitive substrate such as the peroxiredoxin Tpx1 has been selected as a lethality suppressor to favor RNR function at the expense of the non-essential peroxide scavenging function , to allow DNA synthesis and cell growth .
Cysteine residues are not very abundant in proteins , but they are over-represented in functional regions of proteins , such as surfaces and catalytic centers [1] . The thiol group of cysteines is subject of post-translational modifications altering its redox state; several of these oxidation states are reversible , such as sulfenic acid and disulfides . In particular , reversible thiol to disulfide switches happen as a consequence of cellular responses to oxidative stress , and several proteins with reactive cysteine residues undergo oxidations as part of their catalytic cycles ( for a review , see [2] ) . Cells are provided with two major systems meant to control the thiol-disulfide status , the thioredoxin ( Trx ) and the glutaredoxin/glutathione ( Grx/GSH ) systems . Trxs and Grxs catalyze thiol-disulfide exchange reactions , and share a motif known as the Trx fold [3] . Thermodynamically , both types of reductants use as the ultimate electron donor NADPH [4] . Electrons are therefore transferred from NADPH to final substrates through gradients in redox potentials . In the case of Trxs , Trx reductase is the intermediate between NADPH and Trx , while GSH reduces oxidized Grxs , GSH reductase being the link between NADPH and oxidized GSH . Trx was first identified in 1964 as an electron donor for Escherichia coli ribonucleotide reductase ( RNR ) , an enzyme required for DNA synthesis [5] . Grx was later discovered as an alternative electron donor for the same enzyme in E . coli mutants lacking Trx [6] . Many reports indicate that there is cross-talk between both branches of these electron transfer systems and certain redundancy , but it is also clear that there is substrate specificity . From then onwards , it became clear that these oxido-reductases regulate a wide number of processes in eukaryotic and prokaryotic organisms , apart from DNA synthesis and repair , including antioxidant defense and redox regulation , sulfur metabolism or apoptosis; the substrates of Trxs and Grxs mediating these effects are peroxiredoxins ( Prxs ) , GSH peroxidases , methionine sulfoxide reductases , phosphoadenylyl sulfate ( PAPS ) reductase or RNRs ( for reviews on these and other functions of the electron donor cascades , see [2 , 7–12] ) . In most cell types , the only substrate of electron donors which is essential for survival ( and not only for specific cellular processes such as cysteine biosynthesis or oxidative stress tolerance ) is RNR . RNR catalyzes the reduction of ribonucleosides into deoxyribonucleosides , and is therefore essential to provide the building blocks , deoxyribonucleotides ( dNTPs ) , during DNA replication and repair . In eukaryotes , class Ia RNRs are composed of a large subunit , α , containing the catalytic site and two allosteric effector binding sites , that control which substrate is reduced ( specificity site ) as well as the rate of reduction ( activity site ) [13 , 14] , and a small subunit , β , containing a stable diferric-tyrosyl radical cofactor ( oxygen is required for the assembly of the diferric-tyrosyl radical cofactor in RNRs ) , which initiates nucleotide reduction through the transient oxidation of a cysteine to a thiyl radical in the catalytic site of the α subunit . During this process , two local cysteines in the large subunit provide reducing equivalents , and the disulfide bond generated between them , after isomerizing within the same α monomer towards a solvent-exposed position , is reduced by Trx or Grx to yield active RNR . Balanced and sufficient pools of dNTPs have to be present during S phase of the cell cycle , and also to assist in DNA repair . In fact , several studies suggest that a correct supply of dNTPs during DNA replication is important for genome stability and for the prevention of cancer [15 , 16] . Inhibition of RNR activity by the radical scavenger hydroxyurea ( HU ) and other compounds has been used as a chemotherapeutic strategy of numerous cancer types [16 , 17] . RNRs are tightly regulated through many different mechanisms , which include allosteric and oligomeric regulation , transcription of the α and/or β-coding genes to modulate protein levels , inhibition of RNR catalytic activity and regulation of the subcellular localization of the RNR subunits ( for reviews on RNR regulation , see [18 , 19] ) . In Schizosaccharomyces pombe , Cdc22 and Suc22 are the large and small subunits of RNR , respectively [20] . Most studies concerning regulation of fission yeast RNR activity have centered on the RNR inhibitor Spd1 , which affects activity and subunit localization of α and β [21–23] , and on the up-regulation of cdc22 transcription during the S phase and after DNA damage ( for a review , see [19] ) . While the suc22 transcript does not fluctuate with the cell cycle , transcription of cdc22 is up-regulated by the MBF transcription factor , which triggers expression of genes required for the S phase [20 , 24 , 25] . Regarding regulation of cdc22 expression by checkpoint kinases under stress conditions , treatment with HU , which inhibits RNR , decreases the available pool of dNTPs and causes the formation of stalled replication forks and the activation of the DNA replication checkpoint driven by the Rad3 and Cds1 kinases; activated Cds1 phosphorylates and inactivates the Yox1 transcriptional repressor , promoting MBF-dependent cdc22 expression [26] . Regarding the regulation of RNR by cofactors and post-transcriptional modifications , changes in RNR subunit localization in response to iron bioavailability have been recently demonstrated in budding yeast [27] . Nevertheless , very little is known about the redox-dependent cell cycle regulation of RNR activity [28] . In fact , the identity of the S . pombe electron transfer components required for RNR recycling is unknown . Here , we identify Trx1 and Trx3 as the main electron donors of fission yeast RNR , we demonstrate that Trx1 expression is actually up-regulated at S phase at the transcript and protein levels , and that in the absence of Trx1 and Trx3 the DNA replication checkpoint is activated . With the expectation that a complete block of electrons flow should drive to cell lethality by blocking RNR at its oxidized form unless a continuous synthesis of RNR is triggered , we have managed to generate a synthetic lethal combination by deleting the Trx reductase and the Grx1-coding genes . A spontaneous suppressor of this synthetically lethal phenotype is a frame-shift mutation at the beginning of tpx1 , the gene coding for the most abundant consumer of electrons in the cell , the Prx Tpx1 . Our experiments suggest that in the triple knockout strain Δtrr1 Δgrx1 Δtpx1 , the elimination of the main sink of electron favors the reduction of the essential substrate RNR .
S . pombe contains three genes coding for Trxs [29] and one for Trx reductase , trr1 ( Fig 1A ) . Trx1 is the main cytoplasmic Trx [30]; Trx2 is localized to the mitochondria [31]; and Trx3/Txl1 has cytoplasm localization , although it is also associated with the proteasome [32–34] . Trx1 and , to a minor extent , Trx3 are the electron donors of the Prx Tpx1 , essential for aerobic scavenging of peroxides and for signal transduction [35–37] . Thus , while cells lacking Trx1 are extremely sensitive to hydrogen peroxide ( H2O2 ) , Trx2 and Trx3 appear to be dispensable for the defense against oxidative stress [35] . Regarding the other branch of the disulfide reductases , fission yeast expresses only two dithiol Grxs , cytosolic Grx1 and mitochondrial Grx2 , several monothiol Grxs such as Grx4 ( involved in the iron starvation response ) [38–40] , endoplasmic reticulum-located Grx3 [41] and mitochondrial Grx5 ( involved in mitochondrial iron-sulfur cluster assembly ) [42] , and one GSH reductase , Pgr1 ( Fig 1A ) . Pgr1 has been reported to be essential at least during aerobic growth [43] , but cells lacking the reductase can grow under semi-anaerobic conditions . These two cascades have to recycle enzymes which suffer disulfide formation as part of their catalytic functions , such as the essential Cdc22 and the non-essential Tpx1 , Mxr1 or Met16 ( Fig 1A ) . To test the role of these cascades in the turnover of the large RNR subunit , Cdc22 , we combined single or multiple deletion mutations with the expression of a tagged version of the protein , Cdc22-HA . This modification , performed at the chromosomal locus , did not affect cell fitness or cell tolerance to the RNR inhibitor HU ( S1 Fig ) . As shown in Fig 1B , a DTT sensitive , slow migrating band was detected using anti-HA immune-blotting of extracts from asynchronous cultures of wild-type cells expressing Cdc22-HA , corresponding to 10 . 2 ± 1 . 6% of total Cdc22; the sensitivity to the dithiol DTT indicates that the slower migrating band corresponds to a disulfide-containing RNR . In extracts from cells lacking Trx reductase , a band between oxidized and reduced Cdc22 was detected , probably a transient intermediate between RNR and its electron donor . Importantly , cells lacking cytosolic Trx1 , but not the mitochondrial Trx2 , displayed 35 ± 2 . 7% of Cdc22 in its oxidized form . The lack of Trx3 did not significantly affect the amount of Cdc22 disulfide form , but it enhanced the ratio of oxidized-to-reduced form ( 50 ± 3 . 5% ) in a Δtrx1 background . On the contrary , disruption of the Grx branch by deletion of the genes coding for Grx1-to-Grx5 , or GSH reductase , Pgr1 , did not exacerbate disulfide accumulation , unless added to the Δtrx1 Δtrx3 background ( Fig 1C ) . The percentages of Cdc22 oxidation in these and other mutants of the Trx and Grx branches are indicated in Fig 1D . To confirm that Trx deficiencies have an effect on Cdc22 activity , we measured dNTP levels of asynchronous wild-type , Δtrx1 and Δtrx1 Δtrx3 cultures ( Fig 1E for dGTP , and S2A Fig for dATP ) , and detected small but significant decreases in the absence of Trxs . We also measured the percentage of cells at G1 , S and G2 phases in asynchronous cultures from these cells . As shown in Fig 1F , cells lacking Trx1 and , to a larger extent , Trx1 and Trx3 displayed an enlarged population of cells at S phase , indicating that these strains completed DNA replication slower than wild-type cells . We synchronized cultures of cells expressing Cdc22-HA and carrying mutations in several components of the reducing cascades by means of the cdc25-22 allele . Cdc25 is the G2-to-M activating phosphatase of the cyclin-dependent kinase of S . pombe , Cdc2 . Upon shift to the non-permissive temperature , cdc25-22 cells are arrested at the G2/M transition and after dropping the temperature , cells are synchronically released from the arrest . As shown in Fig 2A , Cdc22-HA oxidation cycles in wild-type cells and in cells lacking Trx3 , with a transient peak from 60 to 120 minutes after the release . This peak overlaps with that of the septation index ( Fig 2B ) , which in fission yeast is concomitant with S phase . The peaks of Cdc22-HA oxidation and septation index are delayed and more sustained in cells lacking Trx1 ( or Trx1 and Grx1 ) . Strikingly , in cells devoid of cytosolic Trxs , Δtrx1 Δtrx3 , Cdc22-HA oxidation does not cycle and the protein is maintained at its oxidized form at high levels , around 50–60% . In fact , these cells do not have a clear septation peak . Next , and to confirm that the activity of Cdc22 was compromised in the mutants lacking Trxs , we measured dNTP levels in the previous synchronous cultures . As shown in Fig 2C ( dGTP ) and S2B Fig ( dATP ) , while the levels of dNTPs increased in wild-type cells during cell cycle progression ( 60 and 100 min after release ) , cells lacking Trx1 or Trx1 and Trx3 displayed a reduction of dNTPs levels , pointing that these strains could have compromised DNA synthesis . In fact , when we analyzed the DNA content from the synchronous cultures , we indeed observed a delayed and extended S phase in Δtrx1 cells ( Fig 2D ) . This is even more noticeable in cells in which all the cytosolic Trxs were absent: in Δtrx1 Δtrx3 S phase was not detected by FACS until 120 min after the release , which represents 60 minutes of delay when compared to wild-type cells . Once confirmed that Cdc22 oxidation is exacerbated during catalysis , we tested whether the expression of its main electron donor was also up-regulated during S phase . As shown in Fig 3A , Trx1 protein levels were enhanced at S phase as determined in extracts from block and release experiments . To test whether this protein up-regulation was dependent on a transcriptional event , we used Cyclebase , a repository of published cell cycle experiments [44] , to interrogate genome-wide studies on block and release experiments performed in fission yeast . As shown in Fig 3B , there is a small but consistent cell cycle regulation of trx1 mRNA . The combined peaktime of all published datasets occurs at the beginning of G1 ( Fig 3C ) , a bit later than other S phase transcripts such as cdc22 , cdc18 , yox1 or nrm1 ( S3 Fig ) . All these genes are up-regulated by the MBF transcription factor . To test whether the increase of trx1 mRNA is dependent on MBF , we analyzed its transcript levels upon HU treatment or in cells lacking the MBF repressor Yox1 . As shown in Fig 3D , trx1 transcription does not seem to depend on the MBF complex , contrary to cdc22 . Future work will help us elucidating who triggers the accumulation of trx1 mRNA at G1-to-S transition . Inhibition of RNR activity by HU treatment triggers a DNA replication stress , probably through the decrease in dNTP concentrations and stalling of DNA polymerase at replication forks . In S . pombe , the DNA replication checkpoint is driven by the Rad3 and Cds1 kinases . One target of this cascade is the transcriptional repressor Yox1 , which after phosphorylation by Cds1 is released from the MBF complex and its S phase promoters [26 , 45] ( Fig 4A ) . To test whether RNR inhibition by Trx deficiency can trigger replication stress , we first tested whether cells lacking Trx1 and Trx3 are sensitive to the presence of the RNR inhibitor HU . As shown in Fig 4B , Δtrx1 and Δtrx1 Δtrx3 cells are moderately and severely sensitive to HU , respectively , highlighting their defects in DNA synthesis . In wild-type cells , HU treatment exacerbates the accumulation of total and oxidized Cdc22-HA , as well as of the inhibitory phosphorylation of Yox1 ( Fig 4C ) . Both Trx mutant strains , but specially Δtrx1 Δtrx3 , display constitutive activation of the Rad3-Cds1 checkpoint cascade , as demonstrated by the presence of phosphorylated Yox1 even in the absence of HU stress in this strain background ( Fig 4C ) , and by the enhanced levels of cdc22 mRNA under basal conditions ( Fig 4D ) . We generated a Δtrx1 Δcds1 strain to demonstrate that the weak phosphorylation of Yox1 in Δtrx1 cells is dependent on Cds1 ( Fig 4E ) . Δtrx1 Δtrx3 deletions are synthetic lethal with deletion of cds1 , while a triple Δtrx1 Δtrx3 Δchk1 mutant is viable ( S4 Fig ) ; Chk1 is the effector kinase of the DNA damage checkpoint . We propose that the survival of cells lacking both Trxs depends at least partially on the Cds1-dependent transcriptional up-regulation of the cdc22 gene ( Fig 4A ) . So far , we have demonstrated that Trxs are the main electron donors of Cdc22 , and that cells lacking Trx1 and Trx3 have important defects and activate the replication checkpoint . Taking into account that RNR is an essential protein , we attempted to induce lethality by combining a number of deletions of genes coding for components of the electron pathway cascades ( see Fig 1A ) . Many of the mutants displayed severe growth defects , which could often be rescued by growing the cells in semi-anaerobiosis or in the presence of exogenous GSH ( S1 Table , Fig 5A ) . Indeed , exogenous addition of GSH was sufficient to decrease the ratio of oxidized-to-reduced Cdc22 in mutants lacking Trxs ( Fig 5B and 5C ) and to alleviate some of their growth defects in liquid media ( Fig 5D ) . This suggests that the Grx-GSH branch is a back-up mechanism of reduction of the essential RNR . After exhaustive combination of gene deletions , only two crosses lead to lethality in fission yeast: double deletions of the trr1 ( coding for Trx reductase ) and gcs1 ( codes for glutamate-cysteine ligase , the rate-limiting enzyme on the GSH biosynthetic pathway ) genes , or the double knock-out trr1 and grx1 ( coding for the only dithiol cytosolic Grx ) ( Fig 5E ) . We propose that in this Δtrr1 Δgrx1 strain background RNR would remain oxidized . In spite of the results shown above , and using random spore selection , we unexpectedly obtained a single colony lacking Trr1 and Grx1 and therefore containing a suppressor mutation . To our surprise , we determined by sequence analysis that this mutation laid on the gene encoding the Prx Tpx1 , introducing a one-base deletion at the 26th codon of the open reading frame and subsequently a frame shift and a stop codon at position 71 ( Fig 6A ) . To confirm that the suppressor mutation was linked to loss-of-function of Tpx1 , we performed tetrad analysis to select a triple Δtrr1 Δgrx1 Δtpx1 knock-out strain . As shown in Fig 6B by tetrad dissection , the double Δtrr1 Δgrx1 is synthetic lethal , while tiny colonies of the Δtrr1 Δgrx1 Δtpx1 strain were isolated under semi-anaerobic conditions and could be recovered on plates containing GSH . As shown in S5 Fig , the growth of this triple delete , Δtrr1 Δgrx1 Δtpx1 , displays severe growth defects even under semi-anaerobic conditions , which can be partially overcome by GSH addition . Tpx1 is probably the most demanding substrate of electron donors: there are more than 400 , 000 copies of the protein per cell [46] , and Tpx1 is continuously catalyzing H2O2 detoxification during aerobic growth with the participation of Trx1 , Trx3 and , probably , Grx1 [35 , 36] . The fact that strains such as Δtrx1 Δtrx3 Δgrx1 grow better under semi-anaerobic condition ( Fig 5A ) is an indication that Tpx1 may be competing with Cdc22 for reducing equivalents in cells devoid of the main cytosolic electron donors: the levels of peroxides during semi-anaerobic metabolism are lower than in the presence of oxygen , and therefore Tpx1 is not cycling and demanding electrons to the same extent . To demonstrate that the absence of Tpx1 could positively impinge on the reduced-to-oxidized ratio of RNR , we measured the amount of oxidized and reduced Cdc22-HA in different strains expressing or not Tpx1 . As shown in Fig 6C and 6D , deletion of tpx1 always reduced the percentage of oxidized Cdc22 in three different Trx-deficient strains . We conclude that tpx1 deletion allows the channeling of electrons into the disulfide-bonded RNR , and this is particularly relevant in redoxin mutants . To test whether the competition between Tpx1 and Cdc22 for reducing equivalents could occur in a wild-type background , we forced temporal depletion of reduced Trx1 by Tpx1 during S phase . Exhaustion of reduced Trx1 by Tpx1 can only be accomplished when the Prx is actively scavenging peroxides , but an excess of H2O2 triggers Tpx1 over-oxidation and avoids Trx1 depletion [35] . Therefore , we applied mild oxidative stress in a continuous manner to S phase cultures , by synchronizing wild-type cells expressing Cdc22-HA using the cdc25-22 allele as shown in Fig 2 , and adding or not 100 μM H2O2 at the onset of S phase , with subsequent additions of 25 μM every five minutes , to force Tpx1 oxidation and Trx1-dependent recycling ( Fig 7 ) . Trx1 oxidation was followed in extracts prepared in the presence of 4-acetamido-4′-maleimidylstilbene-2 , 2′-disulfonic acid ( AMS ) as described before [35] . AMS is a bulky thiol alkylating agent: while three moieties of AMS are incorporated in Trx1 when it is in the reduced form , only one AMS is incorporated when Trx1 is oxidized and two of its cysteine residues form a disulfide; slower migrating bands , corresponding to the transient mixed disulfides between Trx1 and its substrates , can also be detected by Western blot upon Trx1 oxidation . As shown in Fig 7A , the S phase-dependent oxidation of Cdc22 does not cause an apparent consumption of reducing equivalents , since the majority of Trx1 remains in the reduced form during the whole cycle . When a continuous addition of mild H2O2 is applied starting at 60 min ( at the onset of S phase ) , a sustained oxidation of Trx1 is accomplished ( Fig 7B ) , which is fully dependent on peroxide scavenging by Tpx1 [35] . Importantly , this Tpx1-dependent depletion of reduced Trx1 enhances the amount of oxidized Cdc22 ( from 12% to 25%; Fig 7C ) , and the disulfide form is maintained for a longer period that in the absence of peroxides ( Fig 7A , 7B and 7C ) . A small but significant cell cycle delay can be observed as a consequence of an elongated S phase , as demonstrated with the septation index ( Fig 7D ) . In conclusion , if oxidative stress emerges during S phase , Tpx1 enzymatic activity jeopardizes the RNR-dependent synthesis of dNTPs through depletion of reduced Trx1 .
Balanced pools of dNTPs have to be accumulated during S phase and after DNA damage and replication stress , and these DNA building blocks are synthesized on demand . In these two scenarios , replication and checkpoint activation , RNR activity is up-regulated through several different mechanisms . We have shown here that the catalytic disulfide formed at the large subunit of RNR , Cdc22 , is reduced by the main cytosolic Trx , Trx1 . Three important conclusions can be extracted from our work: first , the mRNA and protein levels of Trx1 are up-regulated during S phase , what demonstrates a new layer of regulation of RNR . Second , the other cytosolic Trx , Trx3 , may support Trx1 in RNR recycling , so that cells lacking both electron donors suffer from severe replication stress which is partially overcome by the activation of the Rad3-Cds1 checkpoint . Third , the fitness phenotypes of mutants defective in electron donor capacity can be partially alleviated by depletion of another substrate of Trx1 , the Prx Tpx1; elimination of an abundant competitor funnels electrons towards the essential RNR . Regarding the cell cycle-dependent regulation of Trx1 , all the experiments performed so far with synchronized S . pombe cultures highlight the smooth but consistent waves of trx1 transcripts , with a G1 peaktime ( http://cyclebase2 . jensenlab . org/ ) ( Fig 3B ) . We have discarded the participation of the main transcriptional activator of G1-S phase genes , the MBF complex , in trx1 cycling ( Fig 3D ) . Further work will be required to characterize this cell cycle-regulated event . Interestingly , it has recently been reported that colorectal cancer tissues display enhanced protein levels of both RNR and Trx1 , and that inhibition of both proteins simultaneously produced a synergistic anti-proliferation effect in this model [47] . To the best of our knowledge , this is the first report demonstrating that eukaryotic cells carrying Trx deficiencies suffer from replication stress and constitutively trigger the DNA replication checkpoint . In E . coli , an interesting connection between electron donor supplies , activation of DNA replication by DnaA and transcription up-regulation of RNR was proposed by the group of Beckwith [48] . In Saccharomyces cerevisiae , it has been published that mutants lacking Trx1 and Trx2 display a longer S phase [49] , that the total pool of dNTPs from asynchronous cultures is unaffected thanks to the novo synthesized RNR [50] , but that dNTP levels of cells synchronized in S phase are significantly lower than those of wild-type cells [51] . In view of our results , this is probably due to a reduced pool of dNTPs to assist on DNA synthesis [28] . We show here that the checkpoint kinases Rad3 and Cds1 are constitutively active in Δtrx1 Δtrx3 cells , and that this activation is required for the survival of this strain , since the triple Δtrx1 Δtrx3 Δcds1 combination is lethal . We propose that in Δtrx1 Δtrx3 cells the main source of active/reduced Cdc22 is de novo synthesized protein , which arises from the constitutive up-regulation of cdc22 transcription in a Rad3-Cds1-Yox1 dependent manner . The complete lack of Cdc22 recycling should drive cells to lethality . With this idea in mind , we have generated an extensive combination of deletion mutants in most components of the Trx and Grx branches , and one of these combinations resulted in synthetic lethality: Δtrr1 Δgrx1 . The reason why other mutants , such as the quadruple Δtrx1 Δtrx2 Δtrx3 Δgrx1 and Δtrx1 Δtrx3 Δgrx1 Δgrx2 strains , could be isolated under semi-anaerobic conditions but not the aforementioned Δtrr1 Δgrx1 strain is still intriguing to us . It can be speculated that the lack of Trx reductase is more pervasive that the elimination of its substrates due to the accumulation of oxidized Trxs , which may invert their role towards thiol oxidases [52 , 53] , or which may bind to Trx substrates with the same affinity as reduced Trx [54] and block their reduction by other electron donors . A similar result has been reported in other organisms such as E . coli , where lethality or severe sickness can only be accomplished by deletion of the Trx reductase coding gene in combination with a defect in the Grx branch [55] . In S . cerevisiae , deletion of the trx1 , trx2 , grx1 and grx2 genes has been reported to be lethal [56] . Prxs are probably the most demanding cellular substrates of electron donors: they are continuously catalyzing peroxide scavenging during aerobic metabolism , and they are among the most abundant proteins in most cell types . We reported before that Trx1 and , secondarily , Trx3 are the main electron donors of Tpx1 , and cells lacking both cytoplasmic Trxs ( Trx1 and Trx3 ) display constitutively oxidized Tpx1 [35] . Cells lacking Tpx1 cannot grow aerobically on plates; however , strain Δtrx1 Δtrx3 is still viable aerobically , suggesting a secondary role for the Grx/GSH system in Tpx1 reduction . Therefore , Tpx1 and Cdc22 compete for Trx1 , Trx3 and , probably , another component ( s ) of the Grx/GSH cascade . This is not a problem in a wild-type background under most conditions: reduction of Tpx1 by Trx1 is hardly saturated , unless mild oxidative stress is applied , and only for a limited amount of time unless a continuous supply of peroxides is provided ( Fig 7A and 7B ) ; indeed , upon severe H2O2 stress , Tpx1 becomes hyper-oxidized to sulfinic acid and temporarily inactivated [35 , 37 , 57] , which may be beneficial to avoid inhibition of RNR recycling . However , when electron donors become limiting by genetic interventions it is probably advantageous to promote reduction of an essential substrate , RNR , by eliminating a non essential one , a Prx . In fact , other processes improving the fitness of redoxin mutants as well as the oxidized-to-reduced ratio of RNR are semi-anaerobic growth ( by minimizing the activity and electron consumption of Tpx1 in peroxide scavenging; Fig 5A ) and GSH addition ( by providing unlimited reducing power; Fig 5A and S5 Fig ) . In our study , we present evidence for the existence of a novel electron donor for RNR , as the synthetic lethal phenotype of Δtrr1 Δgrx1 mutant can be rescued by eliminating Tpx1 , the major competitor substrate for electrons . It was similarly proposed by Grant and colleagues that PAPS reductase could have an alternative hydrogen donor to Trx1 and Trx2 in budding yeast , since a Δtrx1 Δtrx2 strain grew on minimal media without sulphate under low-aeration growth conditions reducing the generation of reactive oxygen species [56] , and probably minimizing the function of Prxs or GSH peroxidases . We have not identified yet the alternative electron donor of RNR in the Δtrr1 Δgrx1 Δtpx1 background . There are still two or three genes in fission yeast coding for monothiol glutaredoxins ( Grx3 , Grx4 , Grx5 ) , which have been proposed to participate in processes other than disulfide reduction . At least in mammalian RNR , a GSH-mixed disulfide mechanism for Grx-mediated reduction of RNR has been described [58] . Whether S . pombe monothiol Grxs are mediating the channeling of electrons to RNR in a GSH-dependent manner , or whether GSH itself can reduce the disulfide in Cdc22 will have to be elucidated .
Cells were grown in rich medium ( YE ) at 30°C as described previously [59] . When cells were crossed , we chose tetrad dissection or random spore analysis as indicated in the text . For tetrad analysis , asci were dissected by micromanipulation with a Singer Micromanipulator MSM 400 ( Singer Instruments , UK ) . After growth of the dissected spores on YE agar plates under semi-anaerobic conditions , genetic markers were scored by replica-plating on YE-agar plates containing or not the indicated antibiotics , and placing the plates at 30°C under semi-anaerobic conditions in the presence or not of 2 mM GSH , as indicated . Anaerobic liquid cultures were grown in flasks filled to the top with medium at 30°C without shaking . Origins and genotypes of strains used in this study are outlined in Appendix S2 Table , and most of them were constructed by standard genetic methods . A strain with tagged Cdc22-HA , SB104 , was constructed by replacing the cdc22-YFP::kanMX6 cassette of strain AWS16 ( h+ cdc22-YFP::kanMX6 ade6-704 leu1-32 ura4-D18 , kindly provided by A . Carr ) , with a cdc22-HA::natMX6 cassette and by cleaning the auxotrophies . The natMX6 cassette in SB104 was replaced by the hphMX6 cassette , resulting in strain SB110 . Derived strains containing additional deletions were obtained by crossing SB104 or SB110 with the corresponding strains , and plating spores in appropriate media , with the exception of strain AD104 that was obtained by deletion of the pgr1 gene in SB110 . Strains with tagged Yox1-13Myc were obtained by crossing appropriate strains with JA778 ( h- yox1-13myc::kanMX6 ura4-D18 ) or JA779 ( h+ yox1-13myc::kanMX6 ura4-D18 ) . Modified trichloroacetic acid ( TCA ) extracts were prepared blocking thiols with either iodoacetamide or AMS and separated in non-reducing denaturing electrophoresis as previously described [37] . Only when indicated , the reducing agent dithiothreitol ( DTT ) was added to the sample buffer prior to electrophoresis . Cdc22-HA and Yox1-Myc were immuno-detected with monoclonal house-made anti-HA or anti-Myc antibodies , respectively . Trx1 was immuno-detected with anti-Trx1 polyclonal antibody [60] . Anti-Sty1 polyclonal antibody [36] was used as loading control . Relative quantification of protein levels in Western blots was performed by scanning membranes with a Licor 3600 CDigit Blot Scanner ( Licor Inc . , USA ) and using the Image Studio 4 . 0 software . The dATP and dGTP levels were determined by the DNA polymerase-based enzymatic assay as described before [27] . In brief , the incorporation of dATP and dGTP into specific oligonucleotides , containing poly ( AAAT ) and poly ( AAAC ) sequences respectively , by the Klenow DNA polymerase was determined in the presence of excess [3H]-labeled dTTP . We followed a previously published protocol for determining DNA content on isolated nuclei [61] . Briefly , 1x107 cells were fixed in 70% ethanol and nuclei were prepared . Isolated nuclei were treated with RNase A ( 37°C overnight ) and DNA was stained in PBS solution containing 1 μM Sytox green . Temperature-sensitive strains carrying the allele cdc25-22 were cultured in YE at the permissive temperature ( 25°C ) in a shaker water bath until reaching OD600 of 0 . 3 , shifted to the non-permissive temperature ( 36°C ) for 4 hours and then allowed to resume the cell cycle by growing them at 25°C during 3 hours as described [62] . Full arrest at G2/M was checked by microscopy . 5 ml aliquots were taken from non-arrested cells and at different times after release to prepare TCA extracts . Cell cycle progression was monitored with fluorescence microscopy by measuring the septation index of calcofluor-stained cells and by flow cytometry . Total RNA from S . pombe YE cultures was obtained , processed and transferred to membranes as described previously [63] . Membranes were hybridized with the [α-32P]dCTP-labelled cdc22 , trx1 and act1 probes , containing the complete open reading frames . For survival on solid plates , S . pombe strains were grown , diluted and spotted on YE plates containing or not HU at the indicated concentrations and plates were incubated at 30°C for 2–3 days as previously described [26] . To study the survival of strains on solid plates under aerobic or semi-anaerobic conditions , S . pombe strains were grown , diluted and spotted in YE , and plates were incubated at 30°C under aerobic or semi-anaerobic conditions . To grow cells in solid media in an semi-anaerobic environment , we incubated the plates at 30°C in a tightly sealed plastic bag containing a water-activated Anaerocult A sachet ( Merck , Darmstadt , Germany ) [36] , or alternatively in a nitrogen-filled anaerobic chamber ( Forma Anaerobic System , Thermo Electron Corp . ) . When indicated 2 mM GSH was added to YE agar plates . Yeast cells were grown in YE from an initial OD600 of 0 . 2 , with or without the addition of 2 mM GSH , using an assay based on automatic measurements of optical densities , as previously described [64] . | The essential enzyme ribonucleotide reductase ( RNR ) , the rate-limiting enzyme of deoxyribonucleotide synthesis , relies on the thioredoxin and glutaredoxin electron flow cascades for recycling . RNR is tightly regulated in a cell cycle-dependent manner at different levels . Here , we show that cytosolic thioredoxin Trx1 is the primary electron donor for RNR in fission yeast , and that trx1 transcript and protein levels are up-regulated at G1-to-S phase transition . Genetic depletion of thioredoxins triggers the DNA replication checkpoint up-regulating RNR synthesis . Furthermore , deletion of the genes coding for thioredoxin reductase and dithiol glutaredoxin is synthetic lethal , and we show that a loss-of-function mutation at the peroxiredoxin Tpx1-coding gene acts as a genetic suppressor . We propose that in a mutant strain compromised in reducing equivalents , the absence of an abundant and competitive substrate of redoxins , the peroxiredoxin Tpx1 , has been selected as a lethality suppressor to favor channeling of electrons to the essential RNR . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"chemical",
"compounds",
"oxides",
"electron",
"donors",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"dna",
"transcription",
"fungi",
"model",
"organisms",
"dna",
"replication",
"experimental",
"organism",
"systems",
"dna",
"synthesis",
"phase",
"schizosaccharomyces",
"research",
"and",
"analysis",
"methods",
"gene",
"expression",
"schizosaccharomyces",
"pombe",
"chemistry",
"oxidation",
"yeast",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"physical",
"sciences",
"chemical",
"reactions",
"peroxides",
"organisms"
] | 2017 | Lack of a peroxiredoxin suppresses the lethality of cells devoid of electron donors by channelling electrons to oxidized ribonucleotide reductase |
Functional MRI ( fMRI ) experiments rely on precise characterization of the blood oxygen level dependent ( BOLD ) signal . As the spatial resolution of fMRI reaches the sub-millimeter range , the need for quantitative modelling of spatiotemporal properties of this hemodynamic signal has become pressing . Here , we find that a detailed physiologically-based model of spatiotemporal BOLD responses predicts traveling waves with velocities and spatial ranges in empirically observable ranges . Two measurable parameters , related to physiology , characterize these waves: wave velocity and damping rate . To test these predictions , high-resolution fMRI data are acquired from subjects viewing discrete visual stimuli . Predictions and experiment show strong agreement , in particular confirming BOLD waves propagating for at least 5–10 mm across the cortical surface at speeds of 2–12 mm s-1 . These observations enable fundamentally new approaches to fMRI analysis , crucial for fMRI data acquired at high spatial resolution .
Functional magnetic resonance imaging ( fMRI ) experiments have substantially advanced our understanding of the structure and function of the human brain [1] . Hemodynamic responses to neuronal activity are observed experimentally in fMRI data via the blood oxygenation dependent ( BOLD ) signal , which provides a noninvasive measure of neuronal activity . Understanding the mechanisms that drive this BOLD response , combined with detailed characterization of its spatial and temporal properties , is fundamental for accurately inferring the underlying neuronal activity [2] . Such an understanding has clear benefits for many areas of neuroscience , particularly those concerned with detailed functional mapping of the cortex [3] , those using multivariate classifiers that implicitly incorporate the spatial distribution of BOLD [4] , [5] , and those that focus on understanding and modeling spatiotemporal cortical activity [6]–[10] . The temporal properties of the hemodynamic BOLD response have been well characterized by existing physiologically based models , such as the balloon model [11]–[14] . Although the spatial response of BOLD has been characterized experimentally via hemodynamic point spread functions [15]–[18] , it is commonly agreed that the spatial and spatiotemporal properties are relatively poorly understood [19] , [20] . Many studies work from the premise that the hemodynamic BOLD response is space-time separable , i . e . is the product of a temporal HRF and a simple Gaussian spatial kernel . The latter assumed as a simple ansatz or ascribed to diffusive effects , for example [21] . This approach raises the following concerns: ( i ) since the temporal dynamics of the HRF is the focus of most theoretical analyses , e . g . the balloon model , this precludes dynamics that couple space and time , dismissing whole classes of dynamics , such as waves; ( ii ) in practice , employing a static spatial filter then convolving with a temporal HRF on a voxel-wise basis neglects non-separable interactions between neighboring voxels; and ( iii ) calculating temporal correlations between voxels then assumes that the hemodynamic processes responsible for the signal occur on scales smaller than the resolution of the measurements . In summary , neglecting spatial effects such as voxel-voxel interactions and boundary conditions ( e . g . , blood outflow from one voxel must enter neighboring ones ) ignores important phenomena and physical constraints that could be used to increase signal to noise ratios and to improve inferences of neural activity and its spatial structure . These constraints are becoming increasingly relevant , as advances in hardware and software improve the spatial resolution of fMRI by reducing voxel sizes . While treating spatial hemodynamics as a Gaussian is a reasonable first approximation [15]–[20] , [22] , this requires spatiotemporal BOLD dynamics , such as spatially delayed activity to be attributed solely to the underlying neuronal activity , without hemodynamic effects from neighboring tissue , an assumption that may not be valid . In this limit , BOLD measurements would simply impose a spatial low-pass filter of neuronal activity [20] . Several studies have already presented results that challenge this assumption , most strikingly by demonstrating reliable classification of neuronal structures such as ocular dominance or orientation columns [4] , [5] on scales significantly smaller than the resolution of the fMRI protocols used [4] , [5] , [19] , [23] , [24] . Although there are suggestions that the organization of orientation columns may have low spatial frequency components , hemodynamics may also contribute to this effect . Going in the opposite direction , as voxels decrease in size , they must eventually become smaller in linear extent than the hemodynamic response , and thus become highly interdependent . Recent studies [20] , [25]–[27] , have highlighted how BOLD responses involve active changes in cortical vasculature , and hence reflect their mechanical and other spatiotemporal response properties , with spatial scales that are at least partly distinct from the scales of the underlying neuronal activity [20] . Given the above points , the mapping between neuronal activity and the spatially extended BOLD response cannot be assumed to be a spatially local temporal convolution [20] , but should rather be treated in a comprehensive framework that accounts for both spatial and temporal properties and their interactions . A recent theoretical approach [25] treats cortical vessels as pores penetrating the cortical tissue and draws on a rich framework of methods developed in geophysics [28] to derive a physiologically based approximation of the hemodynamics . This model is expressed in terms of a closed set of dynamical equations that reduces to the familiar balloon model [25] in the appropriate limit where spatial effects are averaged over each voxel . It analyzes the spatiotemporal hemodynamic response by modeling the coupled changes in blood flow , pressure , volume , and oxygenation levels that occur in response to neural activity . The objective of the present work is to make and empirically test novel predictions of the model , focusing particularly on spatiotemporal dynamics . We first predict the quantitative spatiotemporal hemodynamic response function ( stHRF ) for physiologically plausible parameters . We find that the model predicts a local response and damped traveling waves , whose speed and range are potentially observable with current high resolution fMRI . Second , we acquire and characterize such high resolution fMRI data from subjects viewing a visual stimulus designed to excite spatially localized neuronal activity in primary visual cortex . We observe hemodynamic waves in these experimental data , whose characteristics confirm our theoretical predictions of wave ranges , damping rates , and speeds , and constrain the physiological parameters of the model .
Cortical hemodynamics and the resulting BOLD signal are modeled by incorporating the physiological properties of cortical vasculature into the theory of fluid flow through a porous elastic medium . The pores are the dense elastic cortical vasculature that penetrate the bulk cortical tissue [25] . In response to a rise in neural activity , local arterial inflow increases , deforming surrounding tissue and thus exerting outward pressure on neighboring tissue . The model predicts coupled dynamical changes in pressure , blood volume , and deoxyhemoglobin ( dHb ) content in the two-dimensional sheet comprising the cortex and its vascular layer ( Figure 1 ) . We refer the reader to the Methods for full details of the model . Here we linearize this model and derive the stHRF , which is the BOLD response due to a spatially point-like , brief increase in neuronal activity . The main stages of this response are as follows: An increase in neuronal activity z ( r , t ) occurs , as a function of time t and position r on the cortical sheet . This causes relaxation of the smooth arterial muscles ( mediated by astrocytes ) , inducing an increased influx of blood ( Figure 1A ) increasing local vessel volume and pressure . Blood then flows to regions of lower pressure , resulting in a redistribution of pressure through the medium ( Figure 1B ) . These pressure changes induce further blood volume changes in adjacent cortical tissue . As a consequence of conservation of mass and momentum , this results in local changes of blood velocity in this adjacent tissue ( Figure 1C ) . As these coupled changes propagate outwards , the viscous properties of blood lead to damping of their amplitude ( Figures 1C , D ) . This is accompanied by outflows to draining veins , reducing vessel volume , and thus yielding further dissipation . Although these two sources of dissipation occur at small scales , they are reflected in the larger scale dynamics of the system . Together , the above processes result in the propagation of pressure changes that travel beyond the ∼1 mm range of direct blood flow between adjacent arterioles and venules . In our theory , the time required for changes in directly coupled adjacent tissue to occur is comparable to the time it takes for the blood to transit the gray matter ( ∼1 s ) . Consequently , these considerations predict propagation speeds of order 1 mm/s . During the change in vessel volume , local deoxyhemoglobin ( dHb ) content also changes ( Figure 1D ) . The influx of arterial blood increases the content of oxygenated hemoglobin ( oHb ) , hence reducing local dHb concentration . Further reductions occur by the removal of dHb by local outflow . As this process occurs , oxygen diffuses passively into the cortical tissue , converting oHb to dHb . Together , these processes result in an initial decrease of local dHb concentration followed by a delayed increase . The BOLD signal thus reflects the net change of blood volume and dHb content . In the case of a spatiotemporally localized neural activation , the predicted BOLD response is given by the stHRF , expressed in Eqs . 1–5 of the Methods . The response to a more general stimulus is obtained by convolving the stimulus with the stHRF , as discussed in Text S1 . Critically , the predicted stHRF ( Eq . 5 ) implies that the hemodynamic response contains a component that propagates as damped traveling waves over spatial scales potentially far greater than those of the neural signal that generated them . In other words , our model predicts that even if neuronal activity is restricted to a very small patch of cortex , it will cause changes in the BOLD signal that propagate for several millimeters over a few seconds . The precise quantitative properties of the predicted BOLD signal depend on several key physiological parameters that can be experimentally determined . Our analysis ( Figure 2 ) indicates that the average vascular stiffness and the rate of damping due to blood viscosity and outflow at boundaries are the most critical parameters ( see Table S1 in Text S1 for a complete list of parameters ) . High stiffness results in a rapid return to equilibrium , thereby increasing the wave speed and range . Conversely high blood viscosity results in strong damping , thereby reducing the range of the waves . Figure 2 shows a range of predicted spatially extended BOLD responses spanning physiologically realistic ranges of these two parameters ( Text S1 ) . A combination of strong damping and low stiffness ( Figure 2 , top left panel ) is predicted to lead to localized responses whose spatial extent is mostly confined to that of the neuronal signal . Although there would still be some weak signal propagation , it is unlikely this would be detectable given typical levels of measurement noise and voxel sizes . This parameter set corresponds to cases where the stHRF spatial scale is smaller than a typical voxel size , where the approach of treating voxels independently would be justified . The opposite extreme of weak damping and high elasticity ( Figure 2 , bottom right ) yields predicted responses that propagate rapidly and far across the cortical sheet . These parameters are unlikely to be relevant experimentally because such extensive waves would likely have already been reported . Between these two extremes there is a broad region of physiologically plausible values of these parameters ( medium damping and/or vascular stiffness ) for which traveling hemodynamic waves are predicted to have properties that are potentially detectable in current experiments but with ranges that would not likely have led to their detection to date . To test the predictions for the theory , high-resolution fMRI data were acquired in primary visual cortex , V1 , from four healthy subjects . The well-known retinotopic mapping of the visual field to V1 allowed us to design simple visual stimuli that resulted in a spatiotemporally localized neural response [29]–[32] . Subjects viewed visual stimuli consisting of three dashed , time-varying ( 4 reversals of the light and dark dashes per second; i . e . , a 2 Hz cycle ) concentric rings at eccentricities of 0 . 6° , 1 . 6° , and 3° ( Figure 3A ) . Subjects were instructed to focus at the center of the screen and perform a simple fixation task ( see Methods ) . These concentric visual stimuli resulted in strong BOLD modulations in early visual cortex , as seen in the example in Figure 3C . As a result of the retinotopic projection from visual field to early cortex ( Figure 3B ) , these concentric rings are projected to lines on the cortical surface , one for each eccentricity . We find that the maximum BOLD modulation occurs on these lines , and that signal modulations weaken away from these peak responses . The three rings were presented at the same time . However , since the hemodynamic responses of the last two rings were not sufficiently separated on the cortex ( in all subjects ) , we focus on the responses to the most inner ring , which was clearly separated ( see Methods ) from the responses of the other rings . The spatiotemporal evoked response is shown in a series of snapshots at different times t with respect to the stimulus onset ( Figure 3C ) . Early responses ( t = 2 . 5 s ) are restricted to the central region ( red ) on these surface patches . As time progresses the BOLD signal near the center rises , and locations successively further from the center demonstrate increasingly delayed rises at t = 2 . 5–7 . 5 s . Finally , outward propagation of positive modulation in the periphery continues , while central responses decrease until they display the well known negative phase of the local post-stimulus undershoot at t = 12 . 5–17 . 6 s . The outward propagation ( Figure 3C ) , discussed in the previous section , suggests the possibility of a traveling wave response , propagating normal to the centerline of the underlying neuronal response . We now focus on characterizing this response . Because stimulation of an isoeccentric curve in the visual field excites an approximately straight line of neurons in V1 [33] , normal directions are clearly identifiable on a flattened cortex . We thus estimate the centerline of the primary isoeccentric response in a flattened representation of V1 ( Figures 4A , B ) , and average the signal change over all points the same distance from this centerline ( Figure 4B ) ( see Methods ) . Repeating this at various distances x and times t reveals the average spatiotemporal response ( Figure 4C ) . Although the signal has been low-pass filtered in the time domain , and averaged along the direction parallel to the stimulus centerline , it is crucial to note that no spatial smoothing has been performed in the x direction . The response has two characteristic spatial scales . Near the center ( |x|<1 mm ) , a local response occurs , with a range similar to that of the expected neural response and whose time variation is similar to that predicted by the ( purely temporal ) balloon model . However , outside this region , the response propagates outward for several mm , with the peak response occurring steadily later as |x| increases ( Figure 4D ) . At |x| = 5 mm the response is delayed , reaching its peak just as the central response |x|<1 mm reverses sign ( Figures 4D , E ) . Furthermore , the amplitude of the propagating response decreases with |x| until it reaches the background level beyond |x| = 5 mm . The above features of the BOLD signal modulations confirm the qualitative theoretical predictions of the spatiotemporal model . The theory also makes quantitative predictions of the waves' propagation speed , range , and damping rate . We next test these predictions and estimate the corresponding parameters through more detailed quantification of the empirical response . To estimate the speed of wave propagation , phase fronts of the BOLD signal were estimated ( see Methods ) . As the hemodynamic disturbance propagates , the spatially dependent time delay is evident ( Figures 3B , 4D ) . This delay also appears as a change in the phase of each frequency component of the response . Analysis of the instantaneous phase , at the frequency of maximum response ( 0 . 1 Hz ) , confirms propagation of phase fronts ( Figure 5A ) . Points on the phase front that intersects the peak of the BOLD activity at x = 0 are overlaid on the spatiotemporal response in Figure 5B . This highlights the different behaviors of the nonpropagating local ( |x|<1 mm ) and propagating ( |x|>1 mm ) components of the BOLD response . As depicted in Figure 5 , our analysis shows that: ( i ) Near the centerline there is a localized response at approximately |x|<1 mm , corresponding to the spatial scale of the expected neural response [34] , including the lateral spreading of thalamocortical projections from the lateral geniculate nucleus to V1 [35] . ( note that all that is required to estimate the properties of the waves , in the following analysis , is that the central region , i . e . Δx , be small compared to the propagation distance of the waves so that we separate the propagating from the local component ) . ( ii ) Propagation occurs away from the center at a roughly constant speed ( constant slope of the phase front ) in both directions . Straight-line fits towards the fovea ( F , red ) and toward the periphery ( P , black ) yield propagation speeds vF = 2 . 3±0 . 2 mm s−1 and vP = 1 . 8±0 . 2 mm s−1 ( s . d . ) , respectively for Subject 1 ( Figure 5B ) . ( iii ) The signal is attenuated as it propagates , with fits to the log-linear plots ( Figure 5C ) yielding spatial damping constants KP = 0 . 33±0 . 02 mm−1 and KF = 0 . 39±0 . 02 mm−1 ( s . d . ) . Characteristic ranges are the reciprocals of these constants; i . e . , about 3 mm for the signal to decrease by a factor of e . Equivalently , the fits vs . t in ( Figure 5D ) imply temporal damping rates ΓP = KPvP = 0 . 56±0 . 04 s−1 and ΓF = KFvF = 0 . 86±0 . 08 s−1 ( s . d . ) . Data of sufficient quality to enable segmentation and primary response identification were obtained in seven of the eight available hemispheres . Data from the left hemisphere of one subject contained signal drop-out and artifact , most likely due to head movement , which prevented artifact-free surface-based reconstruction . Clear evidence of propagating waves in BOLD signal was observed in all seven of these usable data sets ( Figure S5 ) . A total of 14 sets of wave responses were found , from which parameter estimates were able to be made for 12 ( Figure 5E ) . Two cases of propagation toward the periphery in one subject showed interference from the second stimulus ring and were not used . These 12 responses yielded group averages v = 4±2 mm s−1 , K = 0 . 28±0 . 03 mm−1 , and Γ = 0 . 8±0 . 2 s−1 ( s . e . m ) . The scatter plot ( Figure 5E ) shows a correlation ( R2 = 0 . 22 ) between the temporal damping rate and the velocity . As discussed above , the ratio between these two quantities yield secondary estimates for spatial damping and are consistent with the substantially smaller relative uncertainty in the spatial damping constant . Furthermore , this shows that the spatial extent is relatively fixed regardless of the wave properties . This is consistent with the observation that the mean FWHM of the responses is clustered around 4 . 7±0 . 4 mm ( s . e . m ) . To recap , our biophysical theory shows that spatiotemporal hemodynamic responses obey a wave equation , whose key parameters are the wave velocity and a damping constant affecting decay in time and space . These parameters can be estimated from the empirical responses using simple regression analyses . We find that the parameter estimates here all lie within the a priori ranges estimated independently of the model ( Text S1 ) .
The increasing resolution of functional MRI and the development of analysis methods that depend on spatial patterns in these data underline the need for a systematic , quantitative approach to the spatial and temporal properties of the BOLD signal that is based on the properties of the underlying tissue and vasculature . Here we apply a recent physiologically based theory of hemodynamics in cortical tissue and vasculature to derive the linear spatiotemporal hemodynamic response function ( stHRF ) , which is the response to a spatiotemporally localized stimulus of moderate amplitude . High resolution fMRI data are then used to test the predicted BOLD response to localized neuronal modulation in early visual cortex . Just two extra measurable parameters – vβ and Γ - suffice to characterize the spatial properties of the response . The theory used is the first to make a mean-field approximation to cortical vasculature and goes beyond spatially point-like multicompartment models [12] , [36] , [37] as it allows calculation of spatiotemporal hemodynamic responses to general stimuli . It predicts , and the data demonstrate , that the hemodynamic response to a spatially highly localized neuronal drive exhibits traveling waves that propagate over a few mm of cortical tissue . Moreover , the velocity , damping and characteristic range of the observed waves are well within the range of theoretical predictions . These traveling waves have not been previously predicted or reported in human cortex . The central part of the response is non-propagating and has a temporal profile consistent with the standard balloon model [11]–[14] , [38] . A further key implication of the spatial effects in our data is that fMRI voxels should only be treated independently if each voxel is larger than the stHRF scale . For interpreting fMRI acquired at sufficiently high spatial resolution the spatiotemporal properties of the stHRF must also be taken into account . Moreover , when voxels are small , we speculate that propagation of hemodynamic waves beyond their boundaries may underlie the observation that some experiments can be sensitive to structures at scales below the voxel size , including ocular dominance and orientation preference columns [4] , [5] , [24] . The combination of modeling and data allows us to estimate key physiological parameters of the model from observations of individual subjects . This lays the basis for replacing fMRI analysis procedures that rely on purely empirical analysis by ones that relate to the underlying physiology . We have shown how characterization of the spatiotemporal properties of fMRI data allows properties of the cortical tissue and vasculature to be inferred , hence accounting for differences between subjects and , potentially , brain regions . For example , relatively low blood viscosity and/or high tissue stiffness are predicted to lead to longer-range wave propagation . Specific experimental manipulations , such as the use of blood-thinning agents , could be employed to test the predicted changes in wave speed and spatial range . Similarly , the reduction in tissue elasticity that typically occurs with ageing [39] , should be able to be probed noninvasively via its effects on wave velocity , and thereby taken into account when making inferences about neuronal activity in cohorts where age may be a confound . Likewise , regionally specific vascular properties have recently been highlighted as an important potential confound in studies of effective connectivity [40] , [41] , thereby underlining the need for a careful measurement and allowance for hemodynamic effects . It is worth asking why hemodynamic waves have not been previously observed in fMRI . Some reasons are: ( i ) If voxel dimensions are large and sampled over a long time period , the hemodynamic response is not sufficiently resolved to detect propagating waves ( ii ) If the BOLD signal is spatially smoothed , then the spatiotemporal structure of the measured BOLD signal will be averaged out; ( iii ) Wave propagation is confined to occur within the cortical sheet and will be only be readily apparent in surface-based data reconstruction; ( iv ) Hemodynamic waves from a point source ( e . g . , a localized activation in a typical study ) in two spatial dimensions decay more rapidly with distance than from the line source our experiment , where net decay can occur only in the direction perpendicular to the cortical locus of our one-dimensional stimulus . Despite these points , as high resolution protocols and surface-rendered data analysis techniques gain widespread use , the need for quantitative spatial analysis will likewise grow . An important consequence of having hemodynamic traveling waves is that the spatial dynamics of BOLD are not independent of their temporal dynamics . The conventional factorization into spatial and temporal convolution operators is thus not valid in general . A greater understanding of the BOLD response , and brain mapping in general , would come from understanding the spatiotemporal hemodynamic response [20] . The present spatiotemporal HRF provides a solution to this problem , starting from a theory of spatiotemporal hemodynamics . Several other issues arise from having hemodynamic traveling waves . ( i ) The existence of hemodynamic waves mean that spatiotemporal hemodynamics , induced by nearby sources , can interact in a nontrivial way , a property that occurs in the temporal domain , concerning the nonlinear interaction between temporally proximate responses [13] , [38] . ( ii ) These findings cast further doubt on those measures of effective connectivity , such as Granger causality , unless they include a careful treatment of hemodynamic effects [40] , [42] , [43] . ( iii ) On the other hand , experimental designs could exploit the wave properties of hemodynamics by using stimuli that induce resonant properties of cortical tissue - akin to the temporal domain [14]- enhancing detection of the evoked signal . Traveling waves of neuronal or glial origin have been described throughout the brain , including in visual cortex [43]–[46] , raising the question of whether these waves might be responsible for the hemodynamic traveling waves seen in our data . However , several considerations argue against this: ( i ) The close match between the theoretically predicted values and the observed data strongly supports the conclusion that the waves in our data are of hemodynamic origin . ( ii ) Previous studies [45] , [47] , [48] that reported propagating neuronal waves in V1 of similar spatial extent to those seen here demonstrated that these waves are 1–2 orders of magnitude faster ( approximately 200 mm s−1 in cats [47] , 100–250 mm s−1 in primates [45] and 50–70 mm s−1 in rats [48] ) , and waves in cortical white matter travel even faster [43] . Likewise , although the spatial scales may be similar to those presently reported , the diffusion of nitrous oxide - which mediates the coupling between neuronal activity and vasodilation - occurs too rapidly to explain our results [21] . ( iii ) Another possible source of propagating signal of possible relevance are calcium waves traveling via astrocytes because these mediate the neuronal signal in vasodilation . However , these calcium waves travel at ∼10 µm s−1 [49] , which is 2 orders of magnitude slower that the waves reported here . Although hemodynamic waves have not been characterized , detected , or previously modeled , existing work has detailed some spatiotemporal properties of the BOLD response . Previous studies have demonstrated hemodynamic contributions to spatiotemporal BOLD response , including: the effect of draining veins [50] , [51] which induces a latent BOLD signal due to these veins; effects across vascular layers [52] , [53] that induce layer dependent delays of the BOLD response; and general effects of the vascular network [43] that cause delayed BOLD responses across extensive brain regions . Studies have also implemented ways to minimize such effects to improve spatial specificity of functional activations [43] , [50] . The hemodynamic waves are different from the mentioned phenomena in that they exhibit propagation across the cortical surface . As the waves pass through they induce changes in arterioles , capillaries , and venules – not reliant on overall drainage by large veins . The possible interplay of these effects will be subject to future modeling and experimental work . In summary , with advances in imaging technology and data analysis , intervoxel effects will become more pronounced , demanding spatiotemporal analyses based on the underlying brain structure and hemodynamics . By verifying a model that enables such analysis , the present paper opens the way to new fMRI probes of brain activity . These new possibilities include experiments using spatial deconvolution to discriminate between neural and hemodynamic contributions to the spatiotemporal BOLD response evoked by complex sensory stimuli . An important potential application would be to disentangle negative components of the BOLD response from surround inhibition in the visual cortex . Our analysis also affords novel insights and physiological information on neurovasculature , a subject of particular significance to ageing and vascular health . Finally , the combination of the present stHRF with spatially embedded neural field models [8] would allow a systematic and integrated computational framework for inferring dynamic activity in underlying neuronal populations from fMRI data .
The theoretical prediction of the stHRF is derived from a physiologically-based model for spatiotemporal hemodynamics [25] . This model treats brain tissue as a poroelastic medium , with interconnected pores representing the cortical vasculature . The governing equations are a set of nonlinear partial differential equations that connect blood flow velocity v , mass density contributed by blood ( i . e . the part of the total density contributed by blood as opposed to tissue ) ξ , deoxygenated hemoglobin concentration Q , and blood pressure P due to an increase in arterial flow F caused by an increase neural activity z , as a function of time t and position on the cortex r . Although we describe changes in blood volume throughout the text , we model changes in ξ - which is closely related to the fractional volume of blood in tissue: ξ/ρf , where ρf is the density of blood itself . These model equations were recently derived and explained in a separate paper [25] . We provide a synopsis here and apply this with appropriate boundary conditions ( Text S1 ) to calculate the stHRF and derive a hemodynamic wave equation . The dynamics of flow F ( r , t ) are modeled as a damped harmonic oscillator [13] driven by neural activity z ( r , t ) : ( 1 ) The dynamics in Eq . 1 are parameterized by the signal decay rate κ , the flow-dependent elimination constant γ and the resting flow F0 . The neural activity z ( r , t ) drives a distribution of arterial control sites ( see Figure S1 ) , as described further in Text S1 . The vascular response due to this increase in arterial flow is then constrained by physical laws , including conservation laws . Firstly , the conservation of blood mass is embodied by ( 2 ) where cP is a proportionality constant . This conservation law describes how the rate of change of local blood mass density ∂ξ/∂t is determined by the local divergence of the flow ρf∇• ( v ) , the source of mass due to the average inflow of blood F , and the average venous outflow of blood cPP . These latter source/sink terms are mean-field terms that describe the average spatiotemporal hemodynamic processes ( For more details see the Text S1 on the derivation of these terms ) . The rate at which blood travels through the vasculature depends on the elastic response of cortical vessels . This process must conserve momentum , expressed as ( 3 ) where P is the average pore pressure , D parameterizes damping due to blood viscosity , and c1/ρf is the constant of proportionality between pressure gradient and acceleration in the porous medium . This equation describes how forces are directed down pressure gradients , causing blood to accelerate toward regions of lower pressure . These velocity changes are resisted by blood viscosity leading to the resistive term D ( v-vF - vP ) where vF and vP are the blood velocities at inflow and outflow , respectively ( Text S1 ) . The average pressure is related to the elastic properties of blood vessels by the constitutive equation , ( 4 ) where the elasticity of blood vessels is parameterized by the Grubb exponent 1/β ( see Table S1 in Text S1 ) and a proportionality constant c2 , as in previous empirical studies of cerebral blood flow [54] . As changes in local blood volume occur , oxygen diffuses into cortical tissue because of the increased partial pressure of oxygen . This process produces blood deoxyhemoglobin ( dHb ) - whose concentration is represented by Q - from oxygenated hemoglobin . The local concentration of hemoglobin in tissue is a fixed proportion ψ ( in mmol kg−1 ) , of local blood density in tissue , and is thus expressed as ψξ . Hence , the difference ψξ - Q is the amount of oxygenated hemoglobin . If η is the fractional rate at which oxygen passes from oxygenated hemoglobin to cortical tissue , the flow of dHb obeys a conservation equation , similar to Eq . 2 for the conservation of blood mass: ( 5 ) where the difference term ( ψξ - Q ) η on the right hand side introduces the source of dHb , in which dHb is convereted to oHb at a rate η , and the term –QPcP/ ( ξρf ) represents the rate of reduction of dHb concentration due to blood outflow . This assumes that the blood is well mixed so the concentration leaving a vascular unit is Q/ ( ξρf ) at a rate of average venous blood outflow cPP ( Text S1 ) . As in Eq . 2 , the net outflow rate , here dHb ∇• ( Qv ) , is balanced by local changes in content , here ∂Q/∂t . Finally , the measured BOLD signal y is predicted by a recent semi-empirical relation [55] between tissue blood volume content ξ/ρf and dHb content Q to be ( 6 ) where V0 is the resting total volume , Q0 is the resting fraction of dHb , and the constants k1 , k2 , and k3 depend on the acquisition parameters , including field strength and echo time . By restricting the fMRI scans to the occipital pole , we achieved a resolution of 1 . 5×1 . 5×1 . 5 mm3 and 2 s TR echoplanar images ( EPI ) . The stimulus onset was dephased by 250 ms per block for 8 blocks to further increase the effective time resolution . These EPI data were then coregistered to a high-resolution T1-weighted anatomical scan , acquired at 0 . 75×0 . 75×0 . 75 mm3 , so that the spatiotemporal resolution was effectively resampled to 250 ms× ( 0 . 75×0 . 75×0 . 75 ) mm3 . Furthermore , these mappings were restricted to the gray matter by segmenting the anatomical data into gray and white matter . Finally , functional retinotopic scans were used to map out the expected cortical positions in the visual cortex of each subject [3] . Data were acquired on a Philips 3 T Achieva Series MRI machine equipped with Quasar Dual gradient system and an eight-channel head coil . Five healthy subjects ( two female ) ranging from 21 to 30 years participated in this study . The study protocols were approved by ethics boards of the University of New South Wales and Neuroscience Research Australia ( formerly the Prince of Wales Medical Research Institute ) . Participants viewed visual stimuli via a mirror mounted on the head coil at a viewing distance of 1 . 5 m , resulting in a display spanning a diameter of 11° ( or 5 . 5° in eccentricity ) . The visual paradigm was prepared with Presentation® software . Stimulus duration and fMRI pulse timing were logged with 0 . 1 ms accuracy . The stimulus consisted of 3 concentric rings simultaneously presented at 0 . 6° , 1 . 6° , and 3° eccentricity , presented in a block paradigm . ( on for 8 s and off for 12 . 25 s ) . Each annulus was only 1 pixel wide ( roughly 0 . 014° visual angle ) . As known from retinotopic studies , isoeccentric lines in the visual field map to approximately straight lines in primary visual cortex . This stimulus was chosen to exploit this property , optimizing the identification of the primary response and secondary changes in BOLD signal in the orthogonal direction . During the ‘on’ state , the annuli were divided into black and gray dashes that reversed roles 4 times per second ( i . e . , in a 2 Hz cycle ) . During the ‘off’ state , the annuli remained black ( Figure S2 ) . To improve visual fixation , a black fixation cross extended across the entire screen , 4 black circles were permanently present , and a pseudorandomly flickering fixation dot fluctuated between red , green , and blue [56] . Subjects reported that they were able to maintain alertness and attend to the fixation cross throughout data acquisition . Note that off blocks were 12 . 25 s in duration , ensuring that the evoked response was effectively sampled at 250 ms . Each fMRI session consisted of 8 stimulus blocks , consisting of 80+1 fMRI volumes ( the extra scan was due to the delayed onset ) plus an additional 7 ‘off’ scans prior to the first block . Hence each session contained 88 fMRI volumes , a running time of 176 s , 14 such sessions were acquired from each subject . To improve timing accuracy and synchronization we used a monitor refresh rate of 60 Hz for the visual display , and a TR of 2006 ms , rather than 2000 ms , to compensate for system delays . The remaining variability of the stimulus onset precision was logged and used for modeling the experimental design during data analysis . To achieve high resolution , speed , and minimize distortions , we used a SENSE [57] accelerated echoplanar imaging ( EPI ) sequence . Great care was taken to minimize distortion , and each subject's data were carefully investigated to ensure distortion was minimal . Functional data were acquired in 29 1 . 5 mm slices for all but one subject , for whom there were 28 slices , with a 192×192 matrix , 230 mm field of view , and a SENSE factor of 2 . 3 . Functional data were motion corrected and slice scan-time corrected using SPM5 ( SPM software package , http://www . fil . ion . ucl . ac . uk/spm/ ) , then imported into the mrVista- Toolbox ( http://white . stanford . edu/software/ ) for further processing and analysis . The fMRI data were transformed and analyzed in three different spaces: Firstly , the original planar space of the data acquisition , secondly , the 3 dimensional space defined by the high resolution T1 anatomy scans into which the data were aligned , transformed , and spatially up-sampled and finally for a flattened representation of the visual cortex , with maps of phase of the fMRI signals at the left and right occipital pole . Apart from the spatial up-sampling and mapping , no further preprocessing of the data in the spatial dimension was performed . The temporal time series of each voxel were low-pass filtered with a third order Butterworth filter below 0 . 1 Hz . Furthermore , although there were three concentric rings , we focus only on the 0 . 6° ring closest to the fovea when analyzing the spatiotemporal hemodynamic response as this was clearly spatially distinct whereas there was some overlap in responses to the furthest two rings ( 1 . 6° , and 3° ) . Distance measurements were made along the cortical surface using meshes generated by the segmentation on each subject . A shortest path algorithm ( in the VISTA software ) was used to determine these distances on the surface . When a stimulus excites a line of cortex , as in the present case , the hemodynamic response depends only on time and the perpendicular distance x from that line . To analyze these dependences , we estimated the location of the centerline of the primary response on the flattened surface , then measured the average BOLD signal at various distances orthogonal to this response as a function of time since stimulus onset . This was achieved in five steps ( see Text S1 for further details ) : As a hemodynamic disturbance travels , the BOLD signal phase depends on x and t . Phase fronts enable any wave propagation to be tracked at large |x| . To obtain phase estimates from the signal y ( x , t ) , we first constructed the analytic signal [59] , ( 8 ) where ( 9 ) is the temporal Hilbert transform [59] . The phase φ ( x , t ) is then given by ( 10 ) where arg is the complex argument . Maps of this phase are shown in Figure 5A as well as in the Supplementary text for all the data sets . Constant-phase lines represent the phase fronts of the BOLD signal . The empirical estimates for the properties of the wave fronts were calculated from the phase fronts emerging from the peak of the BOLD signal at x = 0 ( Figure 5B ) . From here , the two principal characteristics of the spatiotemporal response can be identified , the local response , close to x = 0 , as a region of near-uniform phase spanning |x|<1 mm , and the propagating component heading away from these regions at |x|>1 mm ( see Figure 5B and Text S1 ) . This is consistent with the expected neural point spread function estimated from independent physiological data [34] . Straight-line fits to this propagating region , as shown in Figure 5B , yielded estimates of the wave velocity νβ . The BOLD signal was measured at each point on the phase front as a function of time ( Figure 5C ) and space ( Figure 5D ) . Transformation to logarithmic scales yielded approximately straight line plots , suggesting exponential decay of BOLD signal in space and time . Linear regression then yielded rate constants of temporal and spatial signal decay . Estimation of the standard error of these linear regressions provided error estimates for these parameters ( see Text S1 ) . | Functional magnetic resonance imaging ( fMRI ) experiments have advanced our understanding of the structure and function of the human brain . Dynamic changes in the flow and concentration of oxygen in blood are observed experimentally in fMRI data via the blood oxygen level dependent ( BOLD ) signal . Since neuronal activity induces this hemodynamic response , the BOLD signal provides a noninvasive measure of neuronal activity . Understanding the mechanisms that drive this BOLD response is fundamental for accurately inferring the underlying neuronal activity . The goal of this study is to systematically predict spatiotemporal hemodynamics from a biophysical model , then test these in a high resolution fMRI study of the visual cortex . Using this theory , we predict and empirically confirm the existence of hemodynamic waves in cortex – a striking and novel finding . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"physics",
"mathematics",
"theoretical",
"biology",
"computational",
"neuroscience",
"interdisciplinary",
"physics",
"signal",
"processing",
"applied",
"mathematics",
"biology",
"computational",
"biology",
"sensory",
"systems",
"biophysics",
"neuroscience",
"neuroimaging",
"engineering"
] | 2012 | Hemodynamic Traveling Waves in Human Visual Cortex |
Viruses that persist despite seemingly effective antiretroviral treatment ( ART ) and can reinitiate infection if treatment is stopped preclude definitive treatment of HIV-1 infected individuals , requiring lifelong ART . Among strategies proposed for targeting these viral reservoirs , the premise of the “shock and kill” strategy is to induce expression of latent proviruses [for example with histone deacetylase inhibitors ( HDACis ) ] resulting in elimination of the affected cells through viral cytolysis or immune clearance mechanisms . Yet , ex vivo studies reported that HDACis have variable efficacy for reactivating latent proviruses , and hinder immune functions . We developed a nonhuman primate model of post-treatment control of SIV through early and prolonged administration of ART and performed in vivo reactivation experiments in controller RMs , evaluating the ability of the HDACi romidepsin ( RMD ) to reactivate SIV and the impact of RMD treatment on SIV-specific T cell responses . Ten RMs were IV-infected with a SIVsmmFTq transmitted-founder infectious molecular clone . Four RMs received conventional ART for >9 months , starting from 65 days post-infection . SIVsmmFTq plasma viremia was robustly controlled to <10 SIV RNA copies/mL with ART , without viral blips . At ART cessation , initial rebound viremia to ~106 copies/mL was followed by a decline to < 10 copies/mL , suggesting effective immune control . Three post-treatment controller RMs received three doses of RMD every 35–50 days , followed by in vivo experimental depletion of CD8+ cells using monoclonal antibody M-T807R1 . RMD was well-tolerated and resulted in a rapid and massive surge in T cell activation , as well as significant virus rebounds ( ~104 copies/ml ) peaking at 5–12 days post-treatment . CD8+ cell depletion resulted in a more robust viral rebound ( 107 copies/ml ) that was controlled upon CD8+ T cell recovery . Our results show that RMD can reactivate SIV in vivo in the setting of post-ART viral control . Comparison of the patterns of virus rebound after RMD administration and CD8+ cell depletion suggested that RMD impact on T cells is only transient and does not irreversibly alter the ability of SIV-specific T cells to control the reactivated virus .
Viral reservoirs are infected cells that persist even in the face of seemingly effective suppressive antiretroviral therapy ( ART ) and can give rise to recrudescent infection when ART is stopped . Reservoir cells include latently infected resting , memory CD4+ T cells , as well as other cells , such as T memory stem cells ( TSCM ) or T follicular helper cells ( Tfh ) [1–8] . Cells harboring latent proviruses carry the virus for the duration of their lifespan . As the half-life of central memory T helper cells is estimated at 44 months [9] , and even longer for the TSCM and Tfh [6 , 10 , 11] , and latently infected cells that do not express viral antigens are invisible to immune clearance mechanisms , such cells can persist for decades , even in patients successfully treated with ART [12–16] . Upon stochastic reactivation , perhaps in connection with homeostatic proliferation or antigen specific stimulation , these quiescent cells can revert their status and start producing new virions [5 , 17] . Even if expression of viral antigens results in immune clearance , the virus will persist as long as proliferation equals or exceeds clearance . ART may suppress most de novo infections of susceptible cells by virions derived from reactivated cells , but viral rebound occurs after variable delays at the cessation of ART , with plasma viral load ( PVLs ) typically rebounding to pretreatment levels [18 , 19] . With an estimated 1 latently infected cell per 1x106 CD4+ T cells [20] , current paradigms predict that the latent viral reservoir is unlikely to be naturally eliminated over the lifetime of an HIV-infected individual on ART [21 , 22] . With the reports of the Berlin patient , the Boston patients and the Mississippi baby , there has been renewed interest in the prospect of achieving viral eradication , or at least sufficient reduction of the reservoir to allow extended viral remission in the absence of continuous ART , both considered as acceptable “HIV cure” strategies [23–26] . Current cure approaches include ART intensification studies [27–31] , infusion of CCR5-gene modified CD4+ T cells following chemotherapy [24 , 32] , enhancement of host HIV-specific immune responses to remove reactivated cells , and variations on the “shock and kill” approach [33–40] . The “shock and kill” strategy has seen the most emphasis . This approach consists of the administration of latency reversing agents ( LRAs ) to induce expression of latent proviruses , with the cells in which virus expression is induced being eliminated by viral cytopathic effects or host immune responses . De novo infection of susceptible cells by LRA-induced virus is prevented by ongoing ART [41 , 42] . This strategy represents a logical approach that theoretically should eventually result in the curbing/elimination of the reservoir . Several types of LRAs have been tested and have shown at least a limited success in activating the latent reservoir , including histone deacetylase inhibitors ( HDACis ) [38 , 43 , 44] , protein kinase C ( PKC ) agonists , such as bryostatin-1 and prostratin [35 , 45] and the bromodomain-containing protein 4 inhibitor JQ1 [34 , 36] . Of these , HDACis have shown the most promise in reactivating the virus from the reservoirs . Valproic acid , givinostat , entinostat , vorinostat ( suberoylanilide hydroxamic acid , SAHA ) , panobinostat and romidepsin ( RMD ) are the most studied HDACis; their effects on reactivating the virus from the reservoir are variable [38 , 42 , 46–48] . RMD has been shown to be among the most potent inducers of HIV in both in vitro and ex vivo models [38] . Recent studies have also documented modest activity of RMD to increase plasma viremia in both HIV-infected patients [49] and SIV-infected macaques [50] on ART . However , RMD has also been reported in vitro to inhibit host CTLs that target infected cells [51] . Elimination of reactivated resting CD4+ T cells in which expression of latent proviruses is induced by LRAs is proving to be more challenging than originally proposed [52] . Initially , it was believed that LRAs capable of potent and extensive induction of latent proviruses could be identified and that upon reactivation , infected cells would be killed by cytopathic effect ( CPE ) and/or host immune responses . Though logical , experience to date has not fulfilled these expectations . Specifically , studies have demonstrated that: only a small fraction of proviruses present in resting CD4+ T cells are reactivated by a single round of HDACi treatment in vitro [53]; reactivation of infected , resting CD4+ T cells by vorinostat did not result in viral CPE-mediated cell death [42]; impairment of host HIV-specific CTLs’ functions is specifically associated with immunodeficiency characteristic to HIV infection even in treated patients [54]; viral clearance by the CTLs only occurs in elite controllers [55]; the effect of various LRAs on latently infected cells , both in vivo and ex vivo , is highly variable [45 , 56 , 57]; in vitro treatment of cells with HDACi can impair CTL function [51]; and , finally , much of the virus present in latently infected cells in individuals that started ART in the chronic phase of infection contains escape mutations for immunodominant epitopes [58] . Therefore , research is currently refocusing on improving the efficiency of viral induction , for example through combinatorial approaches with LRAs targeting different mechanisms that contribute to the establishment and maintenance of viral latency , and boosting the killing of cells expressing induced virus through enhancing host immune responses , particularly the CD8+ CTL responses , by way of therapeutic vaccines , monoclonal antibodies and immune checkpoint inhibitors [39 , 59–62] . Note , however , that complex combinatorial strategies , in which LRAs are combined with immunomodulators , may add difficulty to data interpretation ( as was the case with the Berlin patient [63] ) . In this context , to evaluate different strategies for targeting viral reservoirs , animal models are needed that reproduce key features of HIV infection , while representing experimentally tractable systems that allow a mechanistic evaluation of different interventions . In addition to permitting testing of strategies for which lack of proof of concept or safety concerns may preclude clinical evaluation , animal models , particularly NHP models , offer the opportunity for extensive tissue sampling , to assess what is , after all , a tissue-based disease [64 , 65] . Yet , current animal models have their own limitations: ( i ) For years , SIVmac infection in RMs was more difficult to control than HIV-1 infection , requiring the use of complex and expensive ART regimens [50 , 66–69] , and only recently , a simplified coformulated regimen was reported to effectively control SIVmac infection in RMs [70 , 71]; ( ii ) Humanized mice models , though suitable for addressing some questions relevant to cure research [65 , 72–74] , do not allow detailed assessment of potential virus reservoirs due to limitations on samples obtainable from individual animals and on the feasible durations of treatment . The “Visconti cohort” is a group of French HIV-infected individuals in which prolonged ART was initiated early in infection but eventually halted , and in which a fraction of patients managed to control virus rebound in the absence of continued ART , despite the lack of protective MHC alleles or other known factors that might lead to such an outcome [75] . We developed a NHP model to replicate post-ART control of viral replication in RMs infected with an infectious molecular clone ( SIVsmmFTq ) . This model permits characterization of both the host factors associated with post-treatment control of viral replication and of the dynamics of the viral reservoir in post-treatment controllers . Furthermore , our model can be used to test virus reactivation strategies in the presence of apparently effective immune responses ( in a “shock and effective kill” approach ) . Thus , our model permits the study of LRA efficacy on viral induction without confounding factors such as ART and immunotherapy . We used this model to assess the ability of the HDACi RMD to reactivate virus . We report that RMD can effectively increase virus expression in this model of post-treatment viral control and that RMD administration did not induce a marked or durable alteration of the cellular immune responses in vivo .
To establish a macaque model of post-ART control of virus for studies of approaches to target viral reservoirs that persist in this setting , we identified a SIVsmm strain ( FTq ) [76] that replicates in RMs at levels that are similar to those observed in chronically HIV-1-infected patients . We developed a transmitted/founder ( TF ) infectious molecular clone ( IMC ) of SIVsmmFTq , using the methodology reported in Gnanadurai et al . [77] and used this new IMC to intravenously infect ten Indian RMs . Six RMs were used as controls , in which SIVsmmFTq infection followed its natural course in the absence of any therapeutic intervention , while the remaining RMs served as a study group and received ART , followed by virus reactivation with RMD , as illustrated in Fig 1 . During >1 year follow-up , we show that SIVsmmFTq closely reproduced the patterns of virus replication observed in HIV-1 infection , with high PVL peaks [107−108 viral RNA ( vRNA ) copies/mL] and a robust , but relatively controlled replication during chronic infection ( 104−105 copies/mL ) ( Fig 2a ) . Furthermore , this robust virus replication resulted in a significant depletion of peripheral CD4+ T cells during the acute SIVsmmFTq infection and a partial CD4+ T cell restoration during chronic infection ( Fig 2b ) . The set-point levels of chronic SIVsmmFTq replication being in the range of HIV-1 infection , and lower than those observed with SIVmac239 [78] , we reasoned that , similar to HIV-1 , SIVsmmFTq can be readily controlled with ART . Four RMs intravenously infected with SIVsmmFTq received a combination of tenofovir ( PMPA ) , emtricitabine ( FTC ) and integrase inhibitor ( L-870812 ) for over nine months , starting at 65 days postinfection ( dpi ) . ART resulted in a multiphased decay of plasma viremia with PVLs decreasing to <10 copies/ml in all RMs receiving ART by 30 days on ART , at 95 dpi ( Fig 2a ) . During the follow-up , our priority was to assess the robustness of viral control and assess whether or not viral blips occurred under ART . To this end , RMs on ART were sampled every three days . This very frequent sampling schedule limited the amount of plasma available for viral quantification , preventing us from lowering the detection limit below 10 copies/mL . However , at this detection limit , no detectable vRNA blips were recorded in the plasma over the following 8 months of treatment in any of the RMs receiving ART , confirming our hypothesis that the administered ART regimen successfully controlled SIVsmmFTq replication in these RMs . During treatment , one of the RMs on ART ( RM177 ) died of unrelated conditions ( complications of anesthesia ) , at 112 dpi ( 52 dpt ) . At the time of death , plasma viremia was < 10 copies/mL in this monkey ( Fig 2a ) . The robust control of viral replication in RMs receiving ART impacted the recovery of the CD4+ T cells , which , in spite of being similarly depleted from circulation in both groups of SIVsmmFTq-infected RMs during acute infection , recovered to nearly preinfection levels in the RMs treated with ART ( 9 months of ART , with 8 months of plasma viremia < 10 copies/mL ) . Comparatively , the control RMs showed a more limited restoration at the end of the follow-up ( Fig 2b ) . Furthermore , the fractions of CD4+ and CD8+ T cells expressing immune activation markers were lower in RMs receiving ART compared to controls ( Fig 2c ) . Note , however , that , similar to HIV-infected patients on ART [79 , 80] , and other models of ART-treated SIV-infected RMs [66] , a low level of residual immune activation persisted during antiretroviral therapy in SIVsmmFTq-infected RMs , despite of a robust viral control with ART . These features of the SIVsmmFTq-infected RM model more closely recapitulate key aspects of HIV infection of humans compared to the highly pathogenic SIVmac239 infection . After demonstrating that conventional ART can robustly control SIVsmmFTq replication , we next attempted to reactivate the virus from the reservoir through administration of RMD . One dose of 7 mg/m2 of RMD was administered to the three RMs on ART in a slow perfusion over four hours . Blood samples were obtained during and after RMD treatment to assess the pharmacological effects of the RMD , as well as virus reactivation . We first monitored RMD activity by measuring acetylated histone ( H3 and H4 ) levels in both CD4+ and CD8+ T cells [66] . Histone acetylation increased during the RMD treatment peaked at 6 hours post-RMD treatment initiation and returned to nearly pretreatment levels by 5 days post treatment ( dpt ) , confirming that we had delivered a bioactive dose of the drug ( Fig 3 ) . However , in spite of the documented increase of the levels of acetylated histones , we did not observe any measurable increase in plasma viremia after RMD administration to RMs on ART ( Fig 4 ) . Therefore , one week after completion of RMD treatment , ART was stopped in all RMs . Cessation of ART was followed by rapid and robust rebound of plasma viremia in all three RMs . Viral rebound: ( i ) occurred very rapidly , with SIVsmmFTq being detected in plasma only three days after ART cessation ( Fig 4 ) ; ( ii ) was higher than expected , reaching peak levels of 105−107 copies/ml ( Fig 4 ) , higher than the set-point PVL established during chronic infection , which is often the level of virus rebound observed at the cessation of ART [81]; and ( iii ) was controlled to < 30 copies/ml ( below the limit of detection of our conventional assay ) within 50 days after discontinuation of ART ( Fig 4 ) . We continued to closely monitor PVLs in these RMs using a more sensitive assay and observed that PVLs fluctuated between ≤10 copies/ml and 30 copies/ml , but no animal lost control of the virus over 150 days of observation . Based on the characteristics of the post-treatment dynamics of viremia , these three RMs were labelled as post-treatment controllers [75] ( Fig 4 ) . Due to the unexpected characteristics of the PVL rebound leading to eventual post-treatment control of infection , we designed a new strategy to test whether or not the transient “excess of viral rebound” that followed ART interruption can be attributed to RMD administration . To assess the ability of RMD to induce viral expression in the setting of post-ART spontaneous viral control , we administered three doses of RMD at 35–50 day intervals to the three post-treatment controller RMs . After each treatment , RMD had detectable in vivo activity , as illustrated by increased levels of acetylated histones which peaked at 6 hours postadministration , and returned to nearly pretreatment levels by 5 dpt , as illustrated in Fig 3 . Notably , repeated RMD administration did not result in changes in the levels of acetylated histones , which would have suggested tolerance . As this clear impact on histone acetylation was consistently observed after each RMD administration , we next assessed the ability of RMD to induce increased viral expression . Virus reactivation was monitored by measuring the levels of vRNA in plasma with a single copy PCR assay ( SCA ) specifically developed for SIVsmmFTQ , similar to other previously described assays [82 , 83] . While no detectable rebound of PVLs could be documented in the plasma samples collected at 4 , 6 , 24 and 48 hours post-RMD administration , detectable PVLs were observed starting from 5 dpt in the RMD-treated RMs . Rebounding PVLs peaked at up to 104 vRNA/ml by 13 dpt . After each RMD administration , this consistent virus rebound was gradually controlled to less than 10 copies/ml by 34 dpt ( Fig 5a ) . With repeated RMD administrations , we observed increasingly robust virus rebounds , as documented by increased PVL peaks and longer delays to virus control ( Fig 5a ) . We concluded that RMD has the ability to reactivate the controlled virus in vivo in post-treatment controller RMs . The delays in control of virus rebound , as well as their relative robustness may be due to the fact that , in the absence of ART , new cycles of replication occurred , permitting viral detection in plasma . As such , our study design allowed us to both confirm the efficacy of RMD in reversing latency and document that the virus reactivated after RMD administration is replication-competent . Yet , our results also point to the key observation that the levels of reactivated virus seen with RMD in the presence of ongoing ART are relatively low ( as PVLs were below the SCA limit of detection immediately after RMD administration ) and suggest that only amplification by de novo rounds of infection in the absence of ART allowed us to observe the effect . In the absence of ART , RMD administration did not have a significant impact on the levels of total vDNA from CD4+ memory T cells . Thus , after each round of RMD administration , the levels of vDNA transiently increased . This was due to the study design , which allowed the virus to complete cycles of replication , resulting in the seeding of short-lived memory cells ( i . e . , effector memory cells ) . As these short-lived cells are productively infected and thus rapidly eliminated , the levels of vDNA in memory cells rapidly returned to pretreatment levels between RMD administrations ( Fig 5b ) . To analyze the effect of RMD in more detail , we developed a simple dynamical model of virus production ( see Methods ) , which shows that the slopes of increase of ( the logarithm of ) virus after each cycle of RMD in the absence of ART are related to the enhancement in viral production due to RMD . We estimated the slope of increase in PVLs using a linear-mixed effects model . We found that this slope was not significantly different across the three cycles of RMD treatment in the absence of ART , nor across RMs . The estimated slope was 0 . 418 log10/day ( s . e . 0 . 037 ) . The dynamical model indicates that the increase in viral production over baseline is proportional to this estimated slope ( see Methods ) . Therefore , we can estimate that the increase in viral production due to RMD was between 1% and 5% of the baseline production before RMD , depending on how fast virus is cleared ( 100 day-1 or 20 day-1 , respectively–see Methods for details ) . Assuming that at baseline production is in balance with viral clearance ( P0 = cV0 ) , these percentages allow us to estimate that the average increase in total body virus production attributable to RMD was only between ~150 and ~8000 virions per day . Due to the nature of RMD administration by slow perfusion and the potential complications of prolonged anesthesia , all the animals treated with RMD received fluids throughout the time of drug administration . They also received Boost/Ensure via gavage at the last bleed on day of infusion , then again at 1 , 2 and 3 dpt in order to compensate for the effects of prolonged sedation and extensive bleeding which could have reduced their appetite . In these conditions , we did not observe any major adverse effects of RMD in any RM , with the exception of a slight weight loss ( probably due to frequent anesthesia ) , that recovered by 15 dpt . There were only minimal signs of toxicity after RMD administration , as suggested by the chemistry tests , which were normal at 5 dpt , with the exception of decreases in creatine kinase in all three animals and fluctuations in urea and total protein levels , as illustrated in S1 Fig . Similarly , complete blood counts ( CBCs ) did not show any major change in the blood cell populations indicative of drug toxicity ( S2 Fig ) . Importantly , drug toxicity did not increase with repeated RMD administration . However , due to sample limitations , CBCs and chemistries were only performed prior and 5–7 days after RMD administration and these results should be treated with caution . Therefore , in an additional effort to assess potential RMD related toxicity , we closely monitored samples collected at multiple time points after each of the RMD treatments for levels of plasma lactate dehydrogenase ( LDH ) , a marker of cell injury and death [84] . As illustrated in S3 Fig , RMD administration did not result in a significant increase in the levels of LDH in RMs ( p = 0 . 459 ) ( S3 Fig ) . These results suggest that RMD is effective and safe in RMs at the dose administered in our study . However , in each of the treated RMs , after every RMD administration , a massive , but transient leukopenia was observed , with the overall levels of lymphocytes being reduced by an average of 76% ( range: 55–86% ) ( Fig 6 ) . Leukopenia occurred within 24 hours after RMD administration and lasted less than three days , with the lymphocyte levels being very rapidly restored to pretreatment levels within 5 dpt ( Fig 6a and S2 Fig ) . This pattern was observed in all three RMs , and after every RMD administration , yet the total lymphocyte populations dramatically fluctuated in RM178 with more limited variations in the remaining two RMs ( Fig 6a ) . As a result , both CD4+ and CD8+ T cell counts were drastically reduced upon RMD administration ( Fig 6b and 6c , respectively ) , but similar to the overall lymphopenia , they rebounded to pretreatment levels within one week ( Fig 6b and 6c ) . Since T cell recovery after depletion is typically much slower , this rapid rebound suggests that the apparent reduction in T cell counts observed after RMD administration is not due to real cell depletion . We therefore monitored the CD3 expression on the surface of gated lymphocytes and identified a significant downregulation of CD3 following RMD administration ( Fig 6d ) . The frequency of the CD3+ T cells in the lymphocyte gate decreased after RMD administration , with a concomitant increase in the frequency of CD3-negative cells ( Fig 6d ) . As such , our results suggest that the apparent lymphopenia observed after RMD administration is due to downregulation of lymphocyte surface markers rather than a direct depletion of cells due to drug toxicity . We next monitored levels of activation and proliferation after RMD administration by assessing the fraction of CD4+ and CD8+ T cells expressing the immune activation markers CD25 ( Fig 7a ) , HLA-DR , CD38 ( Fig 7b ) and CD69 ( Fig 7c ) , which increased only transiently in RMs treated with RMD . Increases in the levels of immune activation markers always preceded increases in PVLs suggesting that RMD can activate resting cells ( S4 Fig ) . The fraction of CD4+ and CD8+ T cells expressing the proliferation marker Ki-67 also increased significantly , but this increase tended to be slower than seen for CD69 [85] ( Fig 7d ) . Thus , the fraction of CD4+ and CD8+ T cells expressing Ki-67 peaked at 12 dpt , paralleling viral replication ( Fig 7d ) . The frequency of T cells expressing both Ki-67 and immune activation markers returned to pretreatment levels prior to subsequent RMD administration . Altogether , the dynamics of immune activation and proliferation markers , in combination with findings when the animals were administered RMD while on ART [50] , suggested that these changes were due to RMD administration rather than a response to viral replication , at least in the initial stages after RMD administration . A recent ex vivo study attributed the ability of RMD to reactivate HIV from the reservoir to a major effect exerted by this drug ( similar to other HDACi ) on immune cell effectors , through elimination of CD8+ T cells and a reduction of the cytolytic capabilities of CTLs [51] . As we also observed a major , albeit transient , lymphopenia in RMs after administration of RMD , we next assessed the impact of RMD on SIV-specific T cells in vivo . Functional activity of both CD4+ and CD8+ T cells were monitored by intracellular cytokine staining ( ICS ) measurements of IL-2 , TNF-α , IFN-γ , CD107α and MIP-1β production in response to stimulation with SIVmac239 Gag or Env peptide pools , measured in samples obtained various time points prior to and after RMD administration ( Figs 8 and 9 and S5 and S6 Figs ) . ICS showed that RMD had only a transient impact on the absolute counts of Gag and Env-specific CD4+ and CD8+ T cells ( Fig 10 ) . Thus , while combined cytokine production was transiently hindered after RMD administration , SIV-specific T cell function was rapidly regained , before the viral control was reestablished ( Figs 8 and 9 and S5 and S6 Figs ) . Furthermore , polyfunctionality of the CD4+ and CD8+ T cells was maintained or even boosted after RMD administration , probably as a result of the virus rebound representing a sufficient antigenic stimulus ( Figs 8 and 9 and S5 and S6 Figs ) . The majority of the SIV-specific T cells were positive for the degranulation marker CD107α , which in ICS assays is considered a correlate of cytotoxic potential . Furthermore , the frequency of SIV-specific CD107α positive cells increased after RMD administration ( Figs 8 and 9 and S5 and S6 Figs ) . The same pattern was observed after all RMD treatments . Together with the pattern of viral replication demonstrating control of the rebounding virus after each administration of RMD , our results suggest that in the post-treatment controller RMs that have functional immune responses , the virus reactivated through RMD administration can be effectively cleared by CTLs and that the impact of RMD on SIV-specific T cells is only transient and modest in vivo . Both the dynamics of the immune activation markers and their correlation with PVLs ( Fig 7 and S4 Fig ) , as well as testing of the specific SIV responses , strongly suggested that virus rebound in the RMD-treated RMs was due to virus reactivation after LRA administration and not to loss of viral control through a major ablation of the CTL functions by HDACi . However , to further discriminate between the loss of control and virus reactivation , we modeled in vivo the ablation of CTL responses through direct experimental depletion of CD8+ cells . Post-treatment RM controllers received the M-T807R1 monoclonal antibody ( mAb ) to deplete CD8+ cells , after which plasma viremia , and the number and activation status of CD4+ T cells were compared and contrasted with the results observed after RMD administration . The anti-CD8 mAb successfully depleted peripheral CD8+ cells ( Fig 11a ) and loss of immune control was associated with a dramatic rebound of plasma viremia in all CD8-depleted RMs ( Fig 11b ) . PVLs peaked at up to 107 vRNA copies/ml by 10 days post M-T807R1 administration . As such , the PVLs observed after CD8+ cell depletion were orders of magnitude higher than those observed after RMD administration . PVLs were then slowly controlled over 5 weeks , much slower than after RMD administration , but mirroring the recovery of CD8+ T cells ( Fig 11b ) . This massive viral replication resulted in a significant depletion of the CD4+ T cells in CD8+-depleted post-treatment controller RMs ( Fig 11c ) . CD8+ cell depletion was also associated with a steady increase in the frequency of CD4+ T cells expressing Ki-67 ( Fig 11d ) , which returned to predepletion levels after the rebound of CD8+ cells and control of PVLs . When the levels of CD4+ T cell immune activation and proliferation markers were plotted on the PVLs in the CD8+-depleted post-treatment controller RMs , the viral rebound clearly preceded the increase in the levels of CD4+ T cell immune activation and proliferation markers ( S7 Fig ) . This suggests that the observed virus rebound resulted from the ablation of the immune responses rather than from activation of the reservoir cells , as observed after RMD administration ( S4 and S7 Figs ) . Based on these results clearly documenting different patterns of viral rebound and control after RMD administration and CD8+ T cell depletion , we concluded that RMD administration does not trigger a permanent hindrance on CTL function and that virus rebound after RMD administration is due to the drug administration rather than to an ablation of CTL responses by RMD .
As research for a cure for HIV/AIDS gathers momentum , so does the use of animal models that can be employed to answer multiple key questions related to HIV infection pertinent to potential curative strategies , such as the location and structure of viral reservoirs , the impact of various therapeutic approaches on these reservoirs , as well as the toxicity of candidate LRAs [64] . These questions cannot be addressed without very invasive sampling and without major risks that can be achieved in animal models , but not in a clinical setting where the standard of care for HIV-infected individuals on ART means that in spite of being on chronic medication , they are otherwise able to have a virtually normal life , with a life expectancy that nears that of HIV-uninfected patients [86] . Here , we developed a model of post-treatment virus control of SIV infection that recapitulates features of the human post-treatment controllers [75] . RMs were intravenously infected with a new transmitted founder infectious molecular clone [77] derived from the strain SIVsmmFTq . This strain , which was identified during our previous surveys of SIVsmm diversity in Primate Centers in the US [76 , 87] , has never been passaged in vitro and displays a lower pathogenicity in RMs than the highly adapted SIVmac/SIVsmm strains . We reasoned that since the set-point PVLs of this strain are lower than those of the reference SIVmac strains , SIVsmmFTq may be more readily controlled with ART . At 65 dpi , when the set-point viremia was achieved , but before major immune suppression occurred , RMs received an ART regimen consisting of the NRTIs PMPA and FTC and the integrase inhibitor L-812820 , which is similar to the NRTI/Integrase inhibitor ART regimens containing Raltegravir or Dolutegravir used in combination of antiretrovirals recommended as first line therapy in HIV-infected patients [88] . ART was given continuously for over nine months , which we reasoned should ensure both completion of the first three stages of SIV decay [89 , 90] , as well as a significant decay of the central memory T cells , the major component of the viral reservoir [91] . As shown by our results , our approach was effective , with PVLs controlled to <10 vRNA copies/ml for the duration of treatment , without any blips . Furthermore , at the end of treatment , the biological parameters improved in treated RMs compared to controls , with a trend to better preservation of CD4+ T cells and a partial control of T cell immune activation . Such profiles are characteristic to HIV-infected patients on ART [86 , 91] . We next assessed the usefulness of this new model for testing virus reactivation strategies . For these experiments , we chose RMD ( 94 ) . The rationales for our choice were that HDACi are the most advanced class of LRAs , they are less toxic than other classes of LRAs ( i . e . , PKC agonists or the JQ1 ) [34–36 , 45] , and that RMD is one of the most active HDACi [45 , 49 , 50] . SIVsmmFTq-infected RMs on ART received RMD at a dose of 7 mg/m2 , two-fold higher than the dose previously used in RMs [50 , 92] , but closer to the dose employed in human patients [93 , 94] . During and after treatment , we collected multiple samples and monitored both the effect of RMD on the levels of acetylated histones and the PVLs . We also monitored the side effects of the drug and report that these side effects were minimal , with the exception of dramatic , but transient lymphopenia , the mechanisms of which are currently being investigated in subsequent studies . The tight sampling schedule limited the amounts of plasma available for the SCA , and increased our limit of detection from 1 to 5–10 vRNA copies/ml . However , this is still a very high sensitivity and we could not detect any increase in PVLs after administration of RMD , in agreement with studies in HIV-infected patients , which suggested that RMD has only a limited effect on the reservoir [56] . Therefore , we decided to stop ART seven days after RMD administration . A very rapid and massive virus rebound was observed upon ART interruption in all the SIVsmmFTq-infected RMs . Virus rebound was not unexpected , as in HIV-infected patients , rebound is nearly universal at the cessation of ART . However , detectable levels of SIVsmmFTq were quantifiable in plasma of RMs as early as three days after cessation of ART . This was more rapid than expected , considering that in patients in whom ART is initiated early during infection , similar to our RMs , the average time to detectable virus rebound is ~8 weeks [95 , 96] , and is correlated with the duration on ART [96] , the levels of HIV-1 DNA [97] and with the expression of T-cell exhaustion markers [98] . Even in patients starting ART during the chronic infection and in whom the immune system is exhausted , virus rebound occurs after an average of 2 weeks after ART interruption [99–101] , longer than in the RMs in this present study . Furthermore , while in HIV-infected patients the magnitude of virus rebound at the cessation of ART is generally similar to the set-point levels of viral replication prior to initiation of ART [81] , in our study , the virus rebound was massive ( up to 107 vRNA copies ) , orders of magnitude higher than the set point PVLs established prior to treatment ( i . e . , 104 vRNA copies/ml ) . Therefore , based on the characteristics of the virus rebound , we concluded that the excess of SIVsmmFTq replication observed at the cessation of ART was likely due to RMD , which had been administered only one week before . This was a first indication that RMD successfully contributed to virus reactivation . Unexpectedly , in spite of its massive nature , likely to result in large-scale reseeding of the reservoir , the initial rebound was followed in all RMs by virus control below 50 copies/ml ( the limit of detection of conventional assays ) . Such post-treatment control may raise questions relative to the relevance of our model , considering that ART interruption is associated with permanent loss of virus control in the vast majority of HIV-infected patients [99–101] . Note , however , that there is an important difference between the majority of HIV-infected patients , for whom ART is generally initiated during chronic infection , when they present with a significant immune suppression [99–101] , and our RMs , in which ART was initiated during the initial stages of chronic infection . As such , our model of post-treatment control should be compared with HIV-infected patients in whom ART is initiated early and maintained for prolonged periods of time and for whom post-treatment control was reported to occur [75 , 96 , 102] . The most prominent case of post-treatment control is the Mississippi baby , in whom a very early initiation of ART ( at 30 hours postdelivery ) for a relatively long period of time ( >18 months ) resulted in post-treatment control and delayed virus rebound for 27 months [25] . Similarly , in the ANRS-Visconti cohort of patients , ART was initiated during acute infection for an average of 36 months and post-treatment control was reported to occur in 15% of subjects [75] . Moreover , while clinical trials using short course ART did not report post-treatment control , an impact on the reservoir size has been observed in these studies , resulting in a delayed virus rebound [95 , 96] . Finally , while a <6 month ART regimen initiated early in infection in SIVmac-infected RMs did not result in post-treatment control , a delay of virus rebound was associated with the early ART administration [71] . As such , the overall conclusion from these studies is that a sufficient reduction of the reservoir leading to a delayed virus rebound requires both early initiation and long duration of ART to allow the decay of central memory cells [60] . Here , we fulfilled both these requirements for the post-treatment control , with ART both initiated early in infection and maintained for duration roughly similar to the half-life of memory cells , and , as such , post-treatment control should not be completely unexpected . Also , note that in our newly developed moderately pathogenic SIVsmmFTq model we obtained a more robust control of viral replication with ART than during infection with the highly pathogenic SIVmac239 , in which more aggressive ART regimens are frequently associated with less robust control of the virus and blips of viral replication [50 , 69 , 70] . In this context , it is tempting to speculate that the observed pattern of virus replication at the cessation of ART ( i . e . , massive rebound followed by control ) is due to the fact that early and prolonged administration of ART before the total destruction of the immune system , together with the complete control of a moderately pathogenic virus , contributed to reservoir curbing . An alternative explanation is that preservation of effective cell-mediated immune responses ( through a lower pathogenicity of the virus and an early initiation of ART ) permitted effective control of the virus rebound . Future studies in animals on and off ART will allow us to detail the mechanism ( s ) of the post-treatment control in this model . We next investigated whether or not RMD can reactivate the virus in vivo . We already had indications that RMD could have been at least partly responsible for the excess of viral replication compared to the pre-ART levels observed at treatment interruption . However , at the time of these experiments , the ability of HDACi to reactivate the latent virus and reduce the size of the reservoir was downplayed by ex vivo studies [45] , as well as in vivo studies in humans and macaques on ART [66 , 103] . Therefore , it was critical to design a study which could unequivocally confirm the efficacy of RMD in reactivating the latent virus . There was also a debate in the field regarding the strategy of choice for measuring the reservoir . While authors were arguing that the inducible virus , which would be the main source of virus rebound in patients interrupting ART , can only be assessed by employing the quantitative virus outgrowth assay ( Q-VOA ) [104] , other authors were arguing that Q-VOA underestimates the levels of inducible virus [105] . The major argument was that Q-VOAs were negative in both the Mississippi baby [26] and in the Boston patients [23] , while the virus eventually rebounded in all these patients [23 , 25] . It was also argued that PCR-based methods overestimate the size of the reservoir and do not always correlate with the levels of inducible virus , as they detect defective viral genomes [106 , 107] . Since all these arguments were valid and no gold standard for the assessment of the viral reservoir is yet available , we reasoned that for our study , it would be best to use a conventional method for viral quantification to monitor the effects of RMD in vivo ( i . e . , PVL quantification ) . Such an approach would not only permit us to bypass controversies in the field , but would also allow us to simultaneously assess both the ability of RMD to reactivate the virus , and the replicative abilities of the reactivated virus , thus representing a valuable strategy to monitor LRA efficacy in preclinical in vivo screenings of new LRAs . Our rationale was that without ART , the virus reactivated after RMD administration could reinfect new target cells , replicate and amplify , thus increasing the likelihood of detection with conventional PVL assays . We acknowledge that the main limitation of this strategy is that , in the absence of ART , the reactivated replicating virus will reseed the reservoir and prevent direct assessment of whether or not RMD reduces the reservoir . However , the most important question that we wanted to address was whether or not RMD can reactivate replication-competent virus from the reservoir and , thus , whether it has any future in cure research . We performed three additional RMD administrations to RMs off ART , which resulted in an immediate and transient increase of T cell immune activation , returning to pretreatment levels within 24 hours . This T cell activation was followed by virus rebound in all the RMs treated with RMD . PVLs became detectable at 5 days post-RMD administration and peaked at 103−104 vRNA copies/ml by 13 dpt , thus demonstrating that RMD can indeed activate replication-competent virus from the reservoir . PVLs always followed the increases in the T cell immune activation levels , and therefore we concluded that virus rebound is most likely a result of the reservoir cell activation by RMD and not to loss of viral control through immune cell impairment after RMD administration . Meanwhile , we cannot discount the alternative explanation that the virus rebound might have resulted from the induction of an increased number of target cells , with the implication that RMD-induced SIV transcription may have occurred in only very few cells . In this second scenario , RMD impact on viremia might have been mostly indirect , through homeostatic effects . To address these issues , using a simple model , we estimated that virus reactivation corresponded to between 1% and 5% of the pre-RMD viral production . Note that viral production may be underestimated , because we assumed that the drug has immediate effect , whereas there likely is a delay of at least a couple of hours post-treatment . Accounting for this delay would make the estimated slope of viral increase larger and hence production would also be larger . Our results suggesting a demonstrable , albeit limited , efficacy of RMD in reactivating the latent virus are supported by recent studies in both humans [49] and RMs [50] . To address an alternative explanation for the observed rebound , i . e . , RMD-induced rapid suppression of cytokine production from viable T-cells and selective death of activated T cells , with the net result of impairing the activity of cytotoxic T-lymphocytes , as previously reported [51] , we monitored the percentage of SIV-specific T cells in RMs after RMD administration . We report that , while documenting a reduction in the SIV-specific T cells immediately after RMD administration , we did not observe a substantial long-term impact of RMD on cell-mediated immune responses . However , we observed several interesting features that could help explain the results of the studies reporting the deleterious effects of RMD on CTLs . Immediately after RMD administration , we observed a massive , but very transient , reduction in the T cell counts , with the CD3+ T cell counts recovering in less than 5 days post-RMD administration . Due to the extremely transient nature of this reduction , we reasoned that it does not result from a real depletion of CD3+ T cells , but rather points to the downregulation of the surface markers used for cell counts . An alternative explanation for the observed kinetics of T cells might have been their redistribution to tissues , but in the absence of adequate tissue sampling , we could not assess this alternative . However , we showed that the CD3+ T cell reduction was mirrored by a dramatic increase of the negative cells in the lymphocyte gate . As such , it is possible that the effects of RMD on the cellular immune responses may be artefactual and are likely not the factor behind the observed virus rebound . Finally , to understand whether or not the impact of RMD on cell-mediated immunity contributes to the observed results , we modeled the damage of cellular immunity through experimental depletion of CD8+ cells . RMs received the M-T807R1 mAb and the impact of CD8+ cell depletion was monitored by assessing both the levels of viral replication and those of immune activation . This experiment clearly showed that ablation of CD8+ cells contributed to a massive virus rebound , which was orders of magnitude higher than that observed after RMD . One may argue that CD8+ cell depletion is not comparable with the changes observed following RMD administration which occurs through a different mechanism and only partially impacts the CD4+ and CD8+ T cell populations . Further , the M-T807R1 monoclonal antibody impacts the NK cells , which may also contribute to the observed effects of CD8+ cell depletion . However , our major focus was on comparing the patterns of viral reactivation after RMD and CD8+ cell depletion . We report that , contrasting with RMD administration , in which virus reactivation was a result of increased T cell activation , virus rebound following CD8+ cell depletion preceded the increases in immune activation . Due to these clearly different patterns of virus rebound after RMD or CD8+ cell depletion , we concluded that the observed virus reactivation in our studies is due to RMD and not to impairment of CTL responses . Our study design with RMD in the absence of ART , did not prevent the induced virus from reinfecting susceptible cells , leading to additional cycles of viral replication , reseeding the tissues , and altering the final size of the inducible viral reservoir . Further , due to sample size limitations , we could not perform a thorough characterization of the reservoir changes for the different rounds of RMD treatment . Thus , trying to assess in detail the impact of RMD on the reservoir would be an exercise in futility . However , quantification of the total memory cell-associated vDNA levels revealed that , while the size of the viral reservoir apparently did not significantly change between RMD administrations , a transient increase in the levels of memory cell-associated vDNA occurred after each RMD treatment . This might explain the trend to higher levels of viral reactivation after each additional round of RMD: virus seeding of the short-lived effector memory cell population leads to alterations in the composition of the viral reservoir . For example , we may speculate that , due to virus control during prolonged ART , in the initial viral RMD-induced reactivation , the rebounding virus likely originated nearly exclusively from long-lived resting central memory CD4+ T cells . However , with every round , the reactivated virus infects mostly susceptible cells ( i . e . , activated CD4+ memory cells ) which could contribute to plasma virus in the subsequent rounds . In conclusion , our results demonstrate that RMD may be successfully used to reactivate the latent virus and one may expect that , in the presence of an effective immune response , this intervention may curb the reservoir . Studies of virus reactivation in a background of ART , which are currently ongoing , will enable us to assess whether or not RMD administration can significantly reduce the size of the reservoir . Since the drug efficacy in reactivating the virus is rather modest , combination between different LRA classes or with immune modulators might be a strategy of choice for our attempts to reduce/eliminate the reservoir .
All animals were housed and maintained at the University of Pittsburgh according the standards of the Association for Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) , and experiments were approved by the University of Pittsburgh Institutional Animal Care and Use Committee ( IACUC ) . These studies were covered by the following IACUC protocol: 13011370 . The animals were fed and housed according to regulations set forth by the Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act [108] . All RMs included in this study were socially housed ( paired ) indoors in stainless steel cages , had 12/12 light cycle , were fed twice daily and water was provided ad libitum . A variety of environmental enrichment strategies were employed including housing of animals in pairs , providing toys to be manipulated , and playing entertainment videos in the animal rooms . Furthermore , the animals were observed twice daily and any signs of disease or discomfort were reported to the veterinary staff for evaluation . For sample collection , animals were anesthetized with 10mg/kg ketamine HCI ( Park-Davis , Morris Plains , NJ , USA ) or 0 . 7 mg/kg tiletamine ( HCI ) and zolazepan ( Telazol , Fort Dodge Animal Health , Fort Dodge , IA ) injected intramuscularly . At the end of the study , the animals were sacrificed by intravenous administration of barbituates . Ten Indian RMs were included in the study . They were infected with plasma equivalent to 300 tissue culture infectious doses ( TCID50 ) of SIVsmmFTq [76 , 109] transmitted-founder infectious molecular clone . Clone derivation and preparation was similar to that reported by our group previously [77] . None of the RMs included in this study harbored MHC genotypes associated with control of SIV virus replication ( i . e . A*01 , B*08 , or B*17 ) . Further , RMs were selected to be either homozygous or heterozygous for the TRP allele of Trim5α [110] as the infectious molecular clone was constructed to bypass TFP restriction [110] . Sixty days post-SIVsmmFTq infection , after the resolution of acute infection and establishment of the chronic viral setpoint , ART was initiated in four RMs . ART consisted of the reverse transcriptase inhibitors ( R ) -9- ( 2-phosphonylmethoxypropyl ) adenine ( PMPA; tenofovir; 20mg/kg; gift from Gilead Biosciences ) and β-2’ , 3’-dideoxy-3’-thia-5-fluorocytindine ( FTC; emtricitabine; 50mg/kg; gift from Gilead Biosciences ) by once-daily subcutaneous injection , the integrase inhibitor L-870812 ( 20mg/kg; gift from Merck ) b . i . d . for nine months . PMPA and FTC were administered subcutaneously , while L-870812 was administered orally . During treatment , after all the RMs controlled viral replication , one animal was euthanized due to an unrelated clinical condition ( complications of anesthesia ) . After 9 months of ART , prior to treatment cessation , RMs were treated with the LRA RMD ( Istodax , Celgene Corporation , Summit , NJ ) at a dose of 7 mg/m2 in a slow perfusion over four hours . After cessation of ART , the RM controllers received three additional doses of RMD in similar conditions every 35–50 days . Finally , 42 days after the last RMD administration , RMs received the CD8+ cell-depleting monoclonal antibody M-T807R1 ( NIH Nonhuman Primate Reagent Resource , Boston , MA ) at a dose of 50mg/kg . Animals were closely clinically monitored for physical and physiological changes during all stages of the study . Blood was collected from all RMs as follows: three times prior to infection ( -30 , -15 and 0 dpi ) , biweekly for the first two weeks ( 4 , 7 , 10 , and 14 dpi ) and weekly thereafter . This sampling schedule was designed to monitor viral replication during the acute infection and establishment of the viral set point , at which time point ART was to begin after 5 consecutive similar PVL measurements . During ART treatment , a weekly sampling was designed to monitor for viral blips . Upon cessation of ART , blood was sampled every 3 days to monitor virus rebound . After RMD and anti-CD8 mAb administration , the schedule of blood collection was as follows: 0 , 4 and 6 hours post-treatment , followed by 1 , 2 , 5 , 12 , 14 , 21 , 28 , 35 dpt . Within one hour after blood collection , plasma was harvested and peripheral blood mononuclear cells ( PBMCs ) were separated from the blood using lymphocyte separation media ( LSM , MPBio , Solon , OH ) . Blood chemistries and complete blood counts ( CBCs ) were obtained from Marshfield Laboratories ( Cleveland , OH ) from serum and whole blood , respectively . We monitored the levels of viral replication to assess treatment efficacy as well as the impact of RMD administration on the reservoir virus . Most samples were subject to a quantitative reverse-transcription PCR , as described previously [111] . For samples that achieved <50 copies/ml , a single copy assay ( SCA ) was performed , as described [83 , 112] . Large volumes of plasma ( 5–8 ml ) were pooled and virus pelleted by ultracentrifugation at 170 , 000 x g for 30 min in a Sorvall T1270 rotor . To compensate for increased amounts of nonvirus materials in the plasma that can potentially interfere with accurate quantitation , a known amount of RCAS [83] was added to each sample prior to centrifugation . This serves as an internal control to monitor the overall efficiency of the assay . RNA was isolated as follows: virus pellets were suspended in 100 μl proteinase K for lysing and digestion; 400 μl of GuSCN/glycogen ( glycogen acting as a carrier ) was added and followed by 500 μl of isopropanol to precipitate RNA . The RNA samples were then resuspended in 65 μl Tris-HCl , pH 8 . 0 , DTT , RNasin mixture . cDNA was first prepared in triplicate reaction mixtures for each RNA samples in a 96-well PCR plate . SIVsmmFTq gag standard ( diluted to 1 copy/ml ) , two 7500 copy RCAS aliquots and corresponding water ( negative control ) were added the plate . Reaction mixtures contained 10 μl RNA and 12 μl cocktail [reverse transcriptase plus buffer , Superscript III First-strand synthesis Supermix for qRT-PCR kit ( Invitrogen ) ] . cDNA was then synthesized under the following thermal conditions: 25°C for 10 min , 50°C for 50 min , 85°C for 5 min , and 4°C hold . Following reverse transcription , either SIVsmmFTq primers/probe for Gag ( 200 nM and 100 nM , respectively ) or RCAS primers/probe and 30 μl Taqman Gene Expression Master ( Applied Biosystems , Foster City , CA ) were added to their respective wells . The primer and probe sequences are as follows: SIVFTqF: 5’-AAG TCC AAG AAC ACT GAA TGC ATG-3’; SIVFTqR: 5’-TAT AAT TTG CAT GGC TGC CTG ATG-3; SIVFTqProbe: 5’-/56-FAM/AGC GGA GGT/ZEN/AGT GCC AGG ATT CCA GGC/3IABkFQ/-3’; RCASF: 5’-GTC AAT AGA GAG AGG GAT GGA CAA A-3’; RCASR: 5’-TCC ACA AGT GTA GCA GAG CCC-3’; RCASProbe: 5’-/56-FAM/TGG GTC GGG/ZEN/TGG TCG TGC C/3IABkFQ/-3’ . Real-time PCR and data assimilation were performed utilizing an ABI 7900 HT real-time machine under the following thermal profile: 95°C for 10min to activate the polymerase , followed by 50 cycles of 95°C for 15 seconds , 60°C for 1 min . To monitor the impact of ART on major immune cell populations with emphasis on CD4+ T cell restoration and immune activation , the immune cells were immunophenotyped by flow cytometry . First , a two-step TruCount technique was used to enumerate CD4+ and CD8+ T cells in blood , as previously described [113] . The number of CD45+ cells was quantified using 50 μl of whole blood stained with antibodies in TruCount tubes ( BD Biosciences ) that contained a defined number of fluorescent beads to provide internal calibration . The numbers of CD4+ and CD8+ T cells were then calculated based on the ratio of CD4+ and CD8+ T cells to CD45+ cells in whole blood at the same time point . Whole peripheral blood was stained with fluorescently-labeled antibodies ( all antibodies from BD Bioscience , San Jose , CA , USA unless otherwise noted ) , CD4 ( APC ) , HLA-DR ( PE-Cy7 ) , CD45 ( PerCP ) , CD25 ( PE ) , CD69 ( APC-Cy7 ) , CD20 ( APC-H7 ) , CD8 ( PE-Texas Red ) ( Invitrogen ) and CD38 ( FITC ) ( Stemcell ) . For intracellular staining , cells were fixed , permeabilized and stained for Ki-67 ( PE ) . Flow cytometry acquisitions were performed on an LSR II flow cytometer ( BD Biosciences ) and flow data were analyzed with FlowJo software ( Treestar , Ashland , OR , USA ) . To monitor the dynamics of changes in the virus reservoir due to RMD administration , frozen PBMCs were thawed and CD4+ total memory ( effector and central memory ) cells were sorted using magnetic bead kits and an AutoMACs Pro Separator ( Miltenyi Biotec Cambridge , MA ) . Briefly , CD4+ cells were sorted by staining the PBMCs with the NHP CD4+ T cell isolation kit ( Miltenyi Biotech ) . The sorted cells were then stained with CD95 ( PE ) ( BD Bioscience , San Jose , CA ) , and then with anti-PE microbeads ( Miltenyi Biotech ) for the sorting of total memory cells . Sorted cells were pelleted and dry frozen for DNA extraction . After each sort , we removed 105 cells for purity check . The purity checks were performed as follows: cells from the first sort were stained with the following antibodies: CD3 ( V450 ) , CD8 ( PE-C594 ) , CD4 ( PE ) , CD14 ( FITC ) , CD20 ( APC-H7 ) , CD11c ( APC ) , CD123 ( Pe-Cy7 ) , HLA-DR ( PerCP ) . Cells from the second sort were stained with the following antibodies: CD3 ( V450 ) , CD4 ( APC ) , CD95 ( PE ) , CD28 ( PE-Cy7 ) . Stained cells were analyzed on a LSR II flow cytometer and flow data was analyzed using with FlowJo software . Purity checks confirmed that the purity of the sorted populations was higher than 90% . Total DNA was extracted from the cell pellets using Qiagen DNeasy blood and tissue kit ( Qiagen , Valencia , CA ) . Extracted DNA was then subjected to quantification using the same assays for the plasma vRNA quantification , but omitting the reverse transcription step . Simultaneous quantification of CCR5 was done to normalize sample variability and allow accurate quantification of cell equivalents . The CCR5 primer and probe sequences were: RMCCR5F: 5’- CCA GAA GAG CTG CGA CAT CC—3’; RMCCR5R: 3’- GTT AAG GCT TTT ACT CAT CTC AGA AGC TAA C—3’; RMCCR5Probe: 5’- /56-FAM/TTC CCC TAC/ZEN/AAG AAA CTC TCC CCG GTA AGT A/3IABkFQ—3’ . Primers and probes were ordered from Integrated DNA Technologies ( Integrated DNA Technologies ( IDT ) , Coralville , IA ) , and the Taqman Gene Expression mix was from Applied Biosystems ( Applied Biosystems , Foster City , CA ) . The detection limit of the viral DNA quantification assay was 30 copies/106 cells . Treatment efficacy was determined by measuring the levels of histone acetylation in PBMCs using a flow cytometric assay , on samples collected prior to RMD administration and then at 6 hours , 1 , 3 , 5 dpt . Approximately 2 x 106 freshly isolated PBMCs were surface immunophenotyped for 20 min at room temperature in the dark by using the following flow panel: CD69-brilliant violet ( BV ) 421 ( FN50; Biolegend ) , CD4-V450 ( L200 ) , CD14-BV570 ( M5E2; Biolegend ) , CD8-PE ( SK1 ) , CD28-ECD ( CD28 . 2; Beckman Coulter ) , CD95-PE-Cy5 ( DX2 ) , PD-1-PE-Cy7 ( EH12 . 2h7; Biolegend ) and CD3-APC-Cy7 ( SP34-2 ) . Cells were immediately treated with PhosFlow lyse/fix buffer and incubated for 30 min at 37°C , washed twice , permeabilized with 0 . 4% Triton X-100 buffer ( Sigma ) for 10 min at room temperature in the dark , and washed again . Permeabilized cells were then stained intracellularly for 30 min at 4°C in the dark with the following antibodies: acetylated histone ( recognizes several residues on histones H3 and H4; 3HHH4-2C2; Active Motif ) and Ki-67-Alexa Fluor 647 ( B56 ) . Prior to use , the acetylated histone antibody was FITC labeled by using a Zenon reagent kit ( Invitrogen ) , according to manufacturer’s instructions . After washing , stabilizing fixative was added , and approximately 200 , 000 CD3+ T cells were acquired for each sample by using a BD LSR-II flow cytometer . Population gating was performed using corresponding fluorescence minus one ( FMO ) and untreated negative control samples . Frozen PBMC samples were thawed , counted and treated with the following antibody cocktail: CD107α ( BD ) , CD28 ( BD ) , and CD49d ( BD ) in R10 media . The cells were stimulated by four conditions: SIVmac239 Env peptide pool , SIVmac239 Gag peptide pool , positive control staphylococcal enterotoxin B ( SEB ) , and negative control DMSO . SIVmac239 Env and Gag peptide pools were obtained through the AIDS reagent program , Division of AIDS , NIAID , NIH . The samples were incubated for 2 hours at 37°C . Cells were then treated with Brefeldin A ( Sigma ) and monensin ( Sigma ) for 4 hours at 37°C . Cells were then stained with Blue LIVE/DEAD ( Invitrogen ) , CD4 ( APC ) , CD8α/β ( PE-Texas Red ) and CD3 ( V450 ) for surface and TNFα ( AF700 ) , IFNγ ( FITC ) , IL-2 ( PE ) , and MIP-1β ( PE-Cy7 ) for intracellular . Samples were run on a LSR-II flow cytometer and analyzed with FlowJo software . To investigate if leukopenia observed after RMD administration is due to a real destruction of leukocytes following RMD administration , LDH levels in plasma were quantified by ELISA according to the manufacturer’s protocol ( NeoBioLab , Cambridge , MA , USA ) . Results were expressed in ng/ml plasma and the ranges of detection were 5–100 ng/ml . In healthy human individuals , the LDH levels range from 5 . 6 to 226 ng/mL . Levels of LDH were compared between D0 and mean peak using paired t-test . Differences in the levels of leukocytes and immune activation markers were determined using Mann-Whitney U test . GraphPad Prism 6 ( Graphpad software ) was used for all statistical analysis except for mixed-effects models , which were determined using R . To analyze the impact of RMD on activating cells into viral production we used a simple model , similar to a previously published analysis [114] . We assume that before each cycle of RMD after ART interruption , the PVL is in approximate steady state , as indicated by the data . Virus is produced at a constant rate P and is removed at rate c per virus , such that the change in viral load is described by dVdt=P−cV . Before RMD , the PVL is approximately constant , dV/dt = 0 , and thus P0 = cV0 , where we use the subscript to indicate time 0 of each RMD cycle . Note that at each cycle , the initial viral load ( V0 ) can be different . At each RMD dose , we assume that the production of virus is increased by the recruitment of latent cells into productive infection . We model the early increase of viral production by dVdt= ( P0+PR ) −cV . This is only valid before new cycles of infection ensue , because then viral production ( P0+PR ) is no longer constant . This model can be solved to yield V ( t ) =P0c+PRc ( 1−e−ct ) . By taking the log10 of V ( t ) and expanding the result in a Taylor series to first order , we conclude that early on the PVL changes as log10 ( V ( t ) ) ≈ln ( P0/c ) ln ( 10 ) +cPRP0ln ( 10 ) t=ln ( V0 ) ln ( 10 ) +PRV0ln ( 10 ) t , where “ln” represents the natural logarithm , and we used that P0 = cV0 to obtain the last expression . These expressions show that the log10 of PVL should grow approximately linearly in time early on after RMD treatment , with slope given by the term multiplying time , t . Below we will use both of the above expressions . We fitted a linear mixed effects model to the log10 of PVL between 0 dpt of each cycle of romidepsin treatment and the maximum PVL in each case , using the nlme package of R [115] . We used “Time” and “Cycle” of RMD treatment as fixed factors and RMs as a random factor . We check homogeneity of variance ( plot of residuals ) and normality of error ( Normal qq-plot ) . We found that Cycle was a significant factor , but there was no interaction between Cycle and Time . We also found that the slope of increase over time was not significantly different among the three subjects . Thus , our final mixed-effects linear model that best fitted the initial increase in log10 PVL had a different intercept ( i . e . , estimated initial PVL , V0 ) for each macaque and Cycle , but the same slope ( S ) of increase in all cases , and can be represented by log10 ( VMc ) =V0 , Mc+St , where c = 1 , 2 , 3 represents the RMD cycle number after ART cessation and M corresponds to each RM . The fitting estimated Vc0 , M and S in each case . Putting the two approaches together , the linear mixed-effects model estimates with the dynamical model expressions , we see that S = PR/ ( V0ln ( 10 ) ) . Thus , we can estimate PR , which is the increase in viral production per ml and per unit time due to RMD . We multiply this number by the estimated total blood volume of about 500 ml to obtain the total production . In addition , we see also that S = cPR/ ( P0ln ( 10 ) ) , allowing estimation of PR/P0 , which is the relative increase in production of virus over baseline due to treatment . In this last case , we need to know c for which we use a range of c≈20 day-1 [114] to c≈100 day-1 [116] . | Antiretroviral therapy ( ART ) does not eradicate HIV-1 in infected individuals due to virus persistence in latently infected reservoir cells , despite apparently effective ART . The persistent virus and can rekindle infection when ART is interrupted . The goal of the “shock and kill” viral clearance strategy is to induce expression of latent proviruses and eliminate the infected cells through viral cytolysis or immune clearance mechanisms . Latency reversing agents ( LRAs ) tested to date have been reported to have variable effects , both on virus reactivation and on immune functions . We performed in vivo reactivation experiments in SIV-infected RMs that controlled viral replication after a period of ART to evaluate the ability of the histone deacetylase inhibitor romidepsin ( RMD ) to reactivate SIV and its impact on SIV-specific immune responses . Our results suggest that RMD treatment can increase virus expression in this setting , and that it does not markedly or durably impair the ability of SIV-specific T cells to control viral replication . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"antiviral",
"therapy",
"immune",
"activation",
"immunology",
"microbiology",
"preventive",
"medicine",
"developmental",
"biology",
"antiretroviral",
"therapy",
"molecular",
"development",
"cytotoxic",
"t",
"cells",
"infectious",
"disease",
"control",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"white",
"blood",
"cells",
"animal",
"cells",
"t",
"cells",
"viral",
"replication",
"immune",
"response",
"immune",
"system",
"cell",
"biology",
"immunity",
"virology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] | 2016 | Multi-dose Romidepsin Reactivates Replication Competent SIV in Post-antiretroviral Rhesus Macaque Controllers |
Tsetse flies are obligate blood-feeding insects that transmit African trypanosomes responsible for human sleeping sickness and nagana in livestock . The tsetse salivary proteome contains a highly immunogenic family of the endonuclease-like Tsal proteins . In this study , a recombinant version of Tsal1 ( rTsal1 ) was evaluated in an indirect ELISA to quantify the contact with total Glossina morsitans morsitans saliva , and thus the tsetse fly bite exposure . Mice and pigs were experimentally exposed to different G . m . morsitans exposure regimens , followed by a long-term follow-up of the specific antibody responses against total tsetse fly saliva and rTsal1 . In mice , a single tsetse fly bite was sufficient to induce detectable IgG antibody responses with an estimated half-life of 36–40 days . Specific antibody responses could be detected for more than a year after initial exposure , and a single bite was sufficient to boost anti-saliva immunity . Also , plasmas collected from tsetse-exposed pigs displayed increased anti-rTsal1 and anti-saliva IgG levels that correlated with the exposure intensity . A strong correlation between the detection of anti-rTsal1 and anti-saliva responses was recorded . The ELISA test performance and intra-laboratory repeatability was adequate in the two tested animal models . Cross-reactivity of the mouse IgGs induced by exposure to different Glossina species ( G . m . morsitans , G . pallidipes , G . palpalis gambiensis and G . fuscipes ) and other hematophagous insects ( Stomoxys calcitrans and Tabanus yao ) was evaluated . This study illustrates the potential use of rTsal1 from G . m . morsitans as a sensitive biomarker of exposure to a broad range of Glossina species . We propose that the detection of anti-rTsal1 IgGs could be a promising serological indicator of tsetse fly presence that will be a valuable tool to monitor the impact of tsetse control efforts on the African continent .
Tsetse flies ( Glossina spp . ) are notorious transmitters of trypanosome parasites responsible for Human and Animal African Trypanosomiasis ( HAT and AAT ) . Since 2009 , the annual number of reported cases of HAT has dropped below 10000 ( www . who . int; [1] ) with the prospect and challenge of entering into the elimination phase of HAT in the near future [2] , [3] . Additionally , some 46 million cattle in sub-Saharan Africa are estimated to be at risk of contracting AAT making deep inroads in the socio-economical development of this continent [4] . Beside active HAT case detection and treatment of humans as well as prophylactic and curative treatment of animals with trypanocidal drugs , tsetse vector control represents an important component of trypanosomiasis control , which is mainly based on the use of insecticides through the sequential aerosol spraying technique ( SAT ) , ground spraying , insecticide-treated targets or insecticide-treated animals [reviewed in [5] , [6] , [7]] . After a successful campaign as part of an area-wide integrated pest management on Unguja island ( Zanzibar ) [8] , the sterile insect technique has been added to the vector control arsenal , with ongoing activities in Ethiopia , Senegal and Burkina Faso [9] under the auspices of the Pan African Tsetse and Trypanosomosis Eradication Campaign ( PATTEC ) . However , beside laborious conventional entomological surveys , no sensitive rapid tests are yet available to provide a semi-quantitative measure of the evolution of tsetse fly densities in areas subjected to tsetse control interventions . Indeed , easy-to-use monitoring of tsetse fly exposure on a regular basis would be a highly valuable tool in the follow-up of the efficacy of the applied and/or ongoing tsetse fly control activities . The obligatory blood feeding tsetse flies are the cyclical insect vectors of HAT and a majority of AAT infections are initiated by the bite of an infected tsetse fly . Although it can be assumed that all tsetse fly species could act as vector , a number of Glossina species of the Palpalis group ( e . g . G . palpalis spp . , G . fuscipes spp . , G . tachinoides ) and the Morsitans group ( e . g . G . morsitans spp . , G . pallidipes , G . swynnertoni ) are implicated as major vectors for HAT and animal trypanosomiasis . Given that only a limited percentage of these tsetse flies acquire an infection with trypanosomes , vertebrate hosts living in the tsetse fly belt are predominantly exposed to the bites of uninfected flies . It has been demonstrated for a number of hematophagous insects that salivary proteins induce humoral immune responses that could represent attractive sero-epidemiological markers of exposure ( reviewed in [10] ) . The saliva of Glossina morsitans morsitans tsetse flies was documented to contain over 200 protein constituents [11] from which some are implicated in manipulating the vertebrate hemostatic and inflammatory reactions [12] , [13] , [14] . In G . m . morsitans saliva , the most abundant proteins were shown to be highly immunogenic and to correspond to the 43–45 kDa tsetse salivary gland ( Tsal ) protein family [15] . The physiological role of these proteins remains elusive , but biochemical characterization revealed that they are nucleic acid binding proteins with low endonuclease activity [16] . Immunoglobulin responses to tsetse fly saliva have been detected in humans living in Uganda [15] , Democratic Republic of Congo [17] , [18] and Guinea [19] . Also cattle experimentally exposed to tsetse fly bites displayed elevated levels of anti-saliva antibodies [20] . Immunoblotting studies using the immune plasmas have shown that salivary proteins of several tsetse fly species are recognized by the circulating antibodies [15] , [18] , [19] . The highly abundant Tsal proteins were commonly recognized by the human plasmas and an indirect ELISA using recombinant Tsal1 and Tsal2 proteins as antigens was clearly able to differentiate the tsetse-exposed Ugandan plasmas from control plasmas [15] . Recently , a peptide ( amino acids 18–43 ) derived from the G . m . morsitans adenosine deaminase-related TSGF1 protein was evaluated using a panel of human plasmas from West Africa , revealing that obtained ELISA signals correlated with the anticipated levels of tsetse exposure of the tested populations [21] . Allergic and anaphylactic reactions against tsetse fly bites have also been reported , in which IgE antibodies directed against an Antigen5-related salivary allergen are implicated [22] , [23] , [24] . Efforts to develop a serological test based on a TAg5-derived peptide were not yet successful [21] . We here provide experimental evidence that anti-rTsal1 and anti-G . m . morsitans saliva antibodies can mark exposure of mice and pigs to tsetse flies . Although the anti-tsetse saliva ELISA exhibits a better average test performance , we propose that a serological assay based on the individual recombinant Tsal1 protein could be an alternative to assess the exposure of populations or herds to tsetse fly challenge and hence could be used for tsetse fly epidemiological studies , for prioritizing tsetse fly control , for monitoring and evaluating tsetse fly control schemes and for risk assessment of trypanosome transmission in endemic regions .
The experiments , maintenance and care of animals complied with the guidelines of the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes ( CETS n° 123 ) . Rodent care and experimental procedures were performed under approval from the Animal Ethical Committee of the Institute of Tropical Medicine ( Permit Nrs . PAR013-MC-M-Tryp and PAR014-MC-K-Tryp ) . Breeding and experimental work with tsetse flies was approved by the Scientific Institute Public Health department Biosafety and Biotechnology ( SBB 219 . 2007/1410 ) . Pig experiments were approved by the ITM Animal Ethical Committee ( Permit Nr . PAR-021 ) and the Ethical Committee of the Faculty of Veterinary Medicine of Ghent University ( Permit Nr . EC2010/030 ) and were performed in the Ghent university stables under the supervision of a veterinary doctor . Saliva of Glossina m . morsitans from the ITMA tsetse fly colony was collected as outflow from the salivary glands as described elsewhere [16] . Saliva of Tabanus yao was kindly provided by Prof . Ren Lai ( Kunming Institute of Zoology , Yunnan , China ) , Stomoxys calcitrans saliva was harvested from flies received from Dr . Christopher J . Geden ( United States Department of Agriculture , Gainesville , US ) . Recombinant Tsal1 was purified as described elsewhere from inclusion bodies of IPTG-induced Top10F' Escherichia coli host cells harboring a pQE60:Tsal1 plasmid [15] . Tsal1 was resolubilized in 6M guanidinium hydrochloride , enriched by Ni-NTA affinity chromatography ( Qiagen ) and further purified on a Superdex 200 size exclusion column connected to an Akta Explorer ( GE Healthcare ) in 6 M ureum 50 mM Tris pH 8 . 0 and 600 mM NaCl . Protein concentrations were determined using Nanodrop spectrophotometry and samples were stored in aliquots at −20°C . E . coli soluble extract was prepared as an additive for the porcine ELISA assay diluent as described for other porcine assays [25] . The extract was made from a 1L overnight culture of Top10F' E . coli in Terrific broth . The bacterial pellet was resuspended in 12 ml PBS supplemented with a Complete protease inhibitor cocktail tablet ( Roche ) , followed by 5 rounds of 1 minute sonication . The soluble fraction was harvested as the supernatant after 30 minutes centrifugation at 20 . 000× g . E . coli soluble extract was stored in aliquots at −20°C . Groups of eight female outbred mice ( NMRI , Charles River ) were subjected to different intensities of exposure to G . m . morsitans bites , followed by regular blood sampling and evaluation of the antibody responses in ELISA: ( i ) once exposed to a single fly , ( ii ) once exposed to 10 flies , ( iii ) 3 times per week for 3 weeks exposed to a single fly , ( iv ) 3 times per week for 3 weeks exposed to 10 flies and ( v ) not exposed to tsetse fly bites . One mouse that underwent multiple exposures to a single fly succumbed by day 28 after initiation of exposure . Plasmas were collected over a period of 390 days . After this >1 year non-exposure period , six mice that were exposed to single or multiple ( 10 ) flies were selected based on their physical appearance and behavior and were re-exposed to a single G . m . morsitans tsetse fly followed by weekly plasma collection for up to 42 days after re-challenge . Two mice that were previously exposed to the multiple fly bites succumbed within 4 weeks after the boosting . A Glossina species cross-reactivity study was performed by exposing five mice ( OF1 , Charles River ) per group every 3 days for 6 weeks to 10 flies of either G . m . morsitans , G . p . gambiensis , G . pallidipes ( kindly provided by Peter Takac , Institute of Zoology , Bratislava , Slovakia ) or G . f . fuscipes ( kindly provided by the International Centre of Insect Physiology and Ecology , Mbita Point , Kenya ) . Immune plasma was harvested 10 days after the last exposure . Five non-exposed mice served as negative controls . A cross-reactivity study between salivary antigens of hematophagous insects was performed by intrapinna exposure of 6 OF1 mice per group at 3-weekly interval to decreasing amounts of saliva ( 5 , 2 , 1 and 1 µg ) harvested from Glossina morsitans , Stomoxys calcitrans and Tabanus yao . Plasma samples were collected 10 days after the last immunization . A total of 11 female pigs ( Seghers , Belgium ) , hybrids of Belgian Landrace , Large White and a specific company line were used for experimental exposure to three different G . m . morsitans exposure regimens: ( i ) 5 pigs were weekly exposed for 7 weeks to 30 tsetse flies ( high exposure ) , ( ii ) 5 pigs were exposed 2-weekly for 6 weeks to 3 tsetse flies ( low exposure , 1 pig died at day 7 ) and ( iii ) 1 pig was not exposed to tsetse ( negative control ) . The experimental pigs were 6 weeks old at the start of the experiment following a 4-days acclimatisation period in the university stables ( Faculty of Veterinary Medicine , UGhent ) . For anaesthesia of the pigs , intramuscular injection of midazolam ( 0 . 5 mg/kg ) , ketamine ( 10 mg/kg ) and morphine ( 0 . 1 mg/kg ) was used . Blood was collected every week for 11 successive weeks from the jugular vein using Vacutainer EDTA tubes ( BD ) a few minutes after the pigs had been anesthetised and prior to exposure to the tsetse fly bites . After blood collection , the blood was centrifuged at 3000 rpm for 15 minutes and plasma stored at −20°C . Due to the unexpected death of 1 pig , no tsetse challenge and blood sampling was performed on day 7 for all animals . Due to limited housing capacity in the university stables , only two pigs from the low exposure group could be kept for an additional 2-month period of non-exposure , followed by a re-challenge by the bites of 10 G . m . morsitans flies and a weekly plasma collection over a period of 6 weeks . IgG responses in the exposed animals were analyzed by indirect ELISA against rTsal1 and saliva from different hematophagous insects . For this purpose , polystyrene 96 well plates ( Thermo Scientific NUNC MaxiSorp Surface ) were coated overnight at 4°C with 200 ng antigen ( G . morsitans , S . calcitrans or T . yao saliva or rTsal1 ) per well in 0 . 1 M NaHCO3 ( pH 8 . 3 ) . Plates were overcoated for 1 h with 10% fetal bovine serum ( FBS ) at ambient temperature . Serial half plasma dilutions starting from 1∶100 in assay diluent ( PBS/10%FBS ) were applied for 2 h to antigen and FBS-coated wells . For the analysis of porcine plasma samples , 20% Top10F' E . coli soluble extract was added to the assay diluent to reduce unspecific binding to the antigenic coat . Based on the plasma dilution experiments , a 1∶1600 dilution was chosen for the time course analyses . Specific immunoglobulin detection was achieved using horseradish peroxidase conjugated detection antibodies . For the detection of mouse and porcine IgGs respectively a 1∶1000 diluted rabbit F ( ab ) 2 anti-mouse IgG ( STAR13B , Serotec ) and a rabbit anti-pig IgG conjugate ( A5670 , Sigma ) 1∶4000 diluted in PBS/10%FBS were used . Detection was with TMB substrate ( 3 , 3′ , 5 , 5′-Tetramethylbenzidine , Sigma ) and reactions were stopped by the 1∶3 addition of 1N H2SO4 . Optical density ( O . D . ) was measured using a Multiskan Ascent plate reader ( Thermo ) at a 450 nm wavelength . Antigen-specific responses were expressed as the ΔO . D . between antigen and non-antigen-coated wells . Statistical analyses were performed in SAS Version 9 . 3 ( SAS Institute Inc . , Cary , NC , USA ) . The effect of tsetse fly exposure intensity and boosting on anti-G . m . morsitans saliva and anti-rTsal1 immune responses in mice and pigs was analysed over the entire sampling period based on a mixed model with animal as random effect and challenge , time and their interaction as categorical fixed effects and F-tests were performed at the 5% significance level . Pairwise comparisons were performed using Tukey's multiple comparisons technique to adjust the significance level . Cross-reactivity of immune responses induced by different Glossina species and other hematophagous insects was analysed using a linear fixed effects model using the 1∶100 or 1∶1600 dilution with challenge as categorical fixed effect . Pairwise comparisons were performed with control using Dunnett's multiple comparisons technique to adjust the significance level . Intra-laboratory repeatability of the ELISA tests was assessed by the non-parametric Spearman correlation test . Sensitivity and specificity of the assays were assessed by receiver operating characteristic ( ROC ) curve analysis of the ΔO . D . values of exposed and non-exposed animals , starting from 3 weeks after the initial exposure . The area under the ROC curve ( AUC ) was used as a global index of diagnostic accuracy . Kinetics of the IgG clearance in mice was assessed in five mice from the multiple exposure schemes with sufficient remaining plasma for the different timepoints and using the plasma sample with the highest antibody titer ( set as 100% ) to prepare a standard curve . Percent decrease in antibody titers over time was assessed by a two-phase non-linear regression allowing the estimation of the antibody half-life .
Induction of specific antibody responses was assessed by indirect ELISA in mice following different regimens of exposure to Glossina morsitans morsitans bites . Anti-G . m . morsitans saliva and anti-rTsal1 IgGs were detectable within 7 days after the initial exposure and remained persistently detectable up to 390 days ( Figure 1 ) . The detected IgG titers against both antigens correlated with the different intensities of tsetse fly exposure . Differences in number of flies ( 1 versus 10 flies ) as well as differences in frequency ( single versus weekly exposure over 3 weeks ) were detected by ELISA . Exposure to a single tsetse fly bite was sufficient to induce slightly elevated levels of antibodies against G . m . morsitans saliva and rTsal1 ( Figure 1 ) , although the recorded differences with control plasmas were not significant ( p = 0 . 9843 and p = 0 . 6146 for the anti-rTsal1 and anti-G . m . morsitans saliva tests respectively ) . Single exposure of mice to 10 flies resulted in significantly increased reactivity against G . m . morsitans saliva ( p = 0 . 0160 ) but not against rTsal1 ( p = 0 . 8618 ) . With both antigens , statistically significant differences were recorded considering the entire sampling period between control mice and mice subjected to the repeated exposure to 1 and 10 flies ( p<0 . 0001 ) . Both the anti-rTsal1 and the anti-G . m . morsitans saliva ELISA test were able to differentiate between these two repeated tsetse exposure schemes ( p<0 . 0001 and p = 0 . 0012 respectively ) . Based on a standard curve generated using the plasma sample with the highest IgG titer ( i . e . day 28 , mouse exposed to the most intense biting regimen ) , antibody half-lives in five mice exposed multiple times to a single fly or 10 flies were determined by two-phase decay regressions ( Figure 2 ) . The average half-lives ( T1/22 ) of the anti-saliva and anti-rTsal IgGs in a second phase of decay after a first phase immediately after cessation of tsetse exposure were respectively 36 and 40 days ( Figure 2 ) . There was a significant individual variation within the different exposure groups with up to a 10-fold difference between the strongest and weakest responder . On average , a 3 to 4-fold difference in peak IgG titer was recorded between mice exposed to a low ( multiple bites by a single fly ) and high exposure regimen ( multiple bites by 10 flies ) . The antibodies appeared relatively persistent over an evaluation period of more than 1 year . Following a long period of non-exposure , the bite of a single tsetse fly was sufficient to boost the anti-saliva and anti-rTsal1 IgG titers . This boosting appeared independent of the previous exposure intensity ( Figure 3 ) as statistical analysis over the 6-week sampling period revealed the inability of the saliva and rTsal1-based tests to differentiate mice subjected to a low or high initial exposure regimen ( p = 0 . 4141 and p = 0 . 9609 respectively ) . Antibody titers against both antigens reached a plateau within 7 days after re-challenge , while in naive animals the responses were slightly lower ( p = 0 . 0833 and p = 0 . 1925 for the anti-saliva and anti-rTsal1 IgG levels respectively ) and only reached peak titers within 4 weeks after initial exposure ( Figure 3 ) . A strong correlation between anti-saliva and anti-rTsal1 ELISA results was recorded with a Spearman correlation coefficient r of 0 . 92 . Intra-laboratory repeatability of the anti-saliva and anti-rTsal1 detection assays was excellent ( Spearman r = 0 . 98 ) . A comparison of the diagnostic value of the anti-saliva and rTsal1 indirect ELISA was conducted using ROC curve analysis . Comparison of the area under the curve ( AUC ) , based on the mouse plasmas from the experiment presented in Figure 1 and collected at least 3 weeks after the initial tsetse fly exposure , indicated that the anti-tsetse saliva ELISA has a better test performance than the rTsal1-based assay ( AUC 0 . 94 versus 0 . 82 , Figure S1 ) . Mice were repeatedly exposed to the bites of various Glossina species ( G . m . morsitans , G . p . gambiensis , G . pallidipes and G . f . fuscipes ) . It was noted that the feeding performance of G . fuscipes on mice was inferior to those of the other species , resulting in a high mortality of these flies when maintained on mice . G . m . morsitans saliva and rTsal1 were next used as antigens in indirect ELISA to detect the elicited antibodies . A marked cross-reactivity of the anti-saliva IgGs elicited by three Glossina species ( G . m . morsitans , G . p . gambiensis and G . pallidipes , p<0 . 0001 at the 1∶100 plasma dilution ) and a relatively weaker cross-reactivity of G . f . fuscipes ( p = 0 . 0053 ) was detected with G . m . morsitans saliva as antigen ( Figure 4A&B ) . Similarly , antibodies induced by exposure to G . m . morsitans , G . p . gambiensis and G . pallidipes significantly reacted with rTsal1 ( respectively p<0 . 0001 , p = 0 . 0157 and p = 0 . 0002 ) , but no cross-reaction with rTsal1 was detected in mice exposed to G . f . fuscipes ( p = 0 . 9883 ) . This indicated that a number of antigens are sufficiently conserved in the saliva of several Glossina species to allow multi-species detection of exposure using G . m . morsitans saliva . The rTsal1 also enabled detection of exposure of mice to various tested tsetse fly species , except for G . fuscipes ( Palpalis subgenus ) . Mice were experimentally immunized by 4 intradermal injections with the saliva of stable flies ( Stomoxys calcitrans ) or horse flies ( Tabanus yao ) and compared with control mice and mice that were immunized with tsetse fly saliva following the same immunization protocol . Immunization with Tabanus saliva did not result in IgGs that reacted with rTsal1 and total G . morsitans saliva as coating antigens ( p = 0 . 7882 and p = 0 . 7637 at the 1∶100 plasma dilution , Figure 4D&E ) . However , a slight increase that did not reach statistical significance was observed in responsiveness of the Stomoxys exposed plasmas in the rTsal1 and saliva-based assays ( p = 0 . 0837 and p = 0 . 0639 respectively ) . At a standard 1∶1600 dilution , no cross-reaction of the Stomoxys exposed plasmas with rTsal1 and saliva was detected ( p = 0 . 9920 and p = 0 . 9915 respectively ) . Analysis of the published salivary transcriptome of Stomoxys calcitrans suggested the presence of a Tsal1 homologue from which a truncated sequence was published ( GenBank Accession N°: ACN69159 , [26] ) . Within this region , only 36% identity in amino acid sequence was found , which could explain the slightly elevated anti-rTsal1 reactivity of Stomoxys exposed mice . Pigs were experimentally exposed to a low or a high tsetse fly bite regimen , followed by assessment of the antibody production . As compared to the mouse indirect ELISA , we have modified the porcine assay by including E . coli soluble extract in the sample diluent as described elsewhere [25] to reduce the background that was observed particularly onto the rTsal1 antigen which was produced in a bacterial expression system . Under the used assay conditions , anti-rTsal1 and anti-saliva IgGs were detected from day 21 after the first tsetse exposure onwards ( Figure 5 ) . The specific IgG titers in pigs correlated with the intensity of tsetse exposure . Anti-rTsal1 and anti-tsetse saliva IgGs were elevated in tsetse exposed groups as compared to the non-exposed control pig and pre-immune plasmas , with significant differences for the high exposure group ( p = 0 . 0216 and p = 0 . 0030 respectively ) . The repeated exposure of pigs to 30 flies resulted in higher anti-rTsal1 and anti-tsetse saliva IgG titers as compared to the pigs exposed to a low tsetse challenge by 3 flies ( p = 0 . 0014 and p = 0 . 0024 respectively ) . However , exposure to the low exposure scheme did not result in significantly elevated responses in both the rTsal1 and saliva-based ELISA ( p = 0 . 8844 and p = 0 . 4261 respectively , Figure 5 ) . Boosting of 2 pigs from the low exposure group after a 2-month non-exposure period by the bites of 10 flies resulted in elevated anti-saliva IgG titers but was only weakly detectable using the rTsal1-based ELISA and with higher individual variation ( Figure 6 ) . In general , a good correlation between anti-saliva and anti-rTsal1 ELISA results , obtained for the samples from the tsetse fly exposure experiment presented in Figure 5 , was recorded with a Spearman correlation coefficient r of 0 . 79 . Intra-laboratory repeatability of the anti-saliva and anti-rTsal1 detection assays was adequate ( Spearman r = 0 . 94 ) . Comparison of the area under the curve ( AUC ) for the anti-tsetse saliva and anti-rTsal1 ELISA ROC curve indicated that the assay based on total G . morsitans saliva has a better average performance than the anti-rTsal1 IgG detection test ( AUC 0 . 96 versus 0 . 83 , Figure S2 ) .
Assessment of exposure of populations at risk to the bites of tsetse fly vectors could be an important step towards improved control of both HAT and AAT . Given that novel innovative vector control strategies are being developed ( e . g . miniaturized insecticide-treated targets [27] , the release of sterile male insects [8] , [9] ) and deployed on increasingly large scales on the African continent , monitoring the impact of these new as well as conventional interventions on actual exposure to tsetse fly bites is a logical follow-up . The tsetse fly species implicated in the transmission of trypanosomes are not uniform throughout the African continent as more than 30 tsetse species and subspecies exist that play differential roles in parasite transmission , have preferences for specific biotopes and display specific host feeding preferences . A number of species are strongly implicated in HAT transmission , such as flies of the Morsitans subgenus ( e . g . used in this study: G . morsitans morsitans , G . pallidipes ) and flies of the Palpalis group ( e . g . used in this study: G . palpalis gambiensis , G . fuscipes fuscipes ) . In case of AAT , a large panel of tsetse species and other sympatric hematophagous arthropods are respectively involved in biological and mechanical transmission . Especially stable flies ( Stomoxys sp . ) and horse flies ( Tabanus sp . ) play a role in mechanical transmission of animal trypanosomes such as T . vivax . However , these insects are not able to biologically transmit trypanosomes as the parasite cannot complete its life cycle in these insects . Research on a number of hematophagous arthropod vectors has resulted in the concept of exploiting salivary components as specific biomarkers of exposure [15] , [28] , [29] , [30] , [31] , [32] , [33] . An advantage of this approach is that relatively simple serological tests could provide information on actual exposure to the bites of disease vectors and provide a risk indicator of contracting a vector-transmitted disease without the need of strenuous entomological surveys . Whole salivary gland extracts of sand flies [33] , [34] , [35] , triatomine bugs [30] , [31] and various mosquito species [36] , [37] , [38] can be used to assess biting exposure . It has been suggested that this type of serological approach would enable the detection of very low levels of exposure that could remain undetected by entomological trappings [39] . Also differences in exposure level due to vector control interventions ( e . g . insecticide treated bednets ) could be elucidated on the basis of salivary proteins as immunological probes [29] , [34] , [40] , [41] . While total salivary extracts can be used in various ELISA and immunoblot formats , some studies have moved towards the use of recombinant proteins or peptides which could lead to the development of more standardized immune assays [29] , [42] , [43] , [44] , [45] , [46] , [47] , [48] , [49] . Several strategies could be envisaged either aiming at species-specific or pan-species exposure detection . Studies using various tsetse fly species have shown that salivary components are immunogenic in mice , rabbits , cattle and humans [15] , [17] , [18] , [19] , [20] , [23] . Immunoblotting of salivary proteins separated on 1D or 2D protein gels have highlighted immunogenic proteins corresponding to several protein families including endonuclease ( Tsal ) , adenosine deaminase ( TSGF ) , 5′nucleotidase ( 5′Nuc ) and Antigen 5 ( Ag5 ) related proteins [19] . In addition , the immunogenic nature of G . m . morsitans sgp1 , sgp2 and sgp3 resulted in their identification on the basis of immune screening of a phage cDNA expression library [50] . We have previously observed by Western blot analysis that human plasma samples collected in Tororo ( Uganda , [51] ) where G . fuscipes fuscipes is the predominant tsetse fly , commonly recognized the 43–45 kDa Tsal proteins in G . m . morsitans saliva [15] . This suggested that G . m . morsitans Tsal-based immune screening of these Ugandan samples cross-detected exposure to G . fuscipes fuscipes , responsible for Trypanosoma brucei rhodesiense transmission in that area [52] . Plasma samples from tsetse fly exposed individuals in Guinean HAT foci also displayed strong reactivity against the highly abundant Tsal proteins in G . palpalis gambiensis saliva in immunoblots [19] . As the Tsal protein band is commonly recognized in samples from Glossina-exposed humans [15] , [19] , Tsal proteins could have the potential of a pan-Glossina species exposure marker . However , reactivity with Tsal1 and Tsal2 was also observed with a pool of human control plasmas from Bobo-Dioulasso ( Burkina Faso ) [19] . Based on our current study we anticipate that the inclusion criteria for these urban unexposed plasma donors ( not having traveled outside of the city for at least three months , [21] ) might have been insufficiently stringent to exclude circulating anti-Tsal antibodies . Indeed , tsetse flies are present at sites outside the city and given the high immunogenicity and anticipated long antibody persistence , this could explain the documented positive reactions in the immunoblots . Production of a recombinant version of the Tsal1 protein allowed evaluating its effectiveness for detecting tsetse fly exposure in an antibody-detection ELISA . By experimentally exposing mice and pigs to different tsetse fly biting regimens , the high immunogenicity of total G . m . morsitans saliva and Tsal1 as a major constituent was confirmed . In mice , a single bite was proven sufficient to induce a detectable immune response in naive animals and to boost antibody levels in previously exposed animals . Kinetics of the antibody clearance was assessed in mice over a >1 year period , which revealed a long persistence with an average half-life of 36–40 days . In pigs , the apparent antibody clearance rate was faster and although the time of sampling and the number of experimental animals was limited , the anti-saliva IgG half-life could be estimated to be 15 days ( data not shown ) . This clearance rate was consistent with the fast decline of anti-G . m . submorsitans saliva IgG levels observed in cattle within 10 weeks after cessation of tsetse exposure [20] . Both in mice and pigs , boosting of prior induced anti-saliva immunity was observed within 7 days by the anti-G . m . morsitans saliva and anti-rTsal1 ELISA . However , with the used sample sizes and experimental conditions , both the rTsal1 and saliva-based ELISA were unable to statistically differentiate the control pig from pigs exposed to a low tsetse fly challenge . Interestingly , the boosting of immunity in mice seemed independent of the previous exposure intensity . As such , surveys based on rTsal1 or total G . m . morsitans saliva in tsetse fly control intervention zones would give information on the actual tsetse exposure intensity rather than history and could provide a tool to monitor re-invasion . Consistent with this , serological results for cattle in South-West Burkina Faso based on G . palpalis gambiensis whole saliva extract related to the seasonality which directly impacts the intensity of host/vector contact [20] . Also humans living in HAT endemic areas display significant differences in anti-G . p . gambiensis saliva IgGs depending on the study site [19] . Similar results were obtained with a peptide derived from the adenosine deaminase-related protein TSGF1 , although obtained ELISA signals seemed very low [21] . In both the cattle and human ELISA tests , large inter-individual differences were observed [15] , [19] , [20] , [21] . It should be noted that in our study the individual variation in antibody titers in the different exposure groups also varied significantly . In controlled experimental conditions with outbred mice , we recorded up to 10-fold differences in anti-saliva and anti-rTsal1 antibody concentrations between the strongest and weakest responders . Consequently , using the tsetse salivary antigens as immunological probes has the limitation of inherent variability due to individual differences in immunological responsiveness . Nevertheless , we observed a clear overall difference in anti-rTsal1 and anti-saliva IgG levels between the different exposure groups of mice and pigs , with an average 3–4 fold difference in specific antibody concentration between a low exposure ( repeated exposure to 1 fly for mice or to 3 flies for pigs ) and a high exposure regimen ( repeated exposure to 10 flies for mice or to 30 flies for pigs ) . This is in line with observations made for cattle that were experimentally exposed to different biting intensities [20] . Beside population-level studies , individual serological follow-up of selected animals or sentinel animals using the recombinant Tsal1 or the total G . morsitans saliva could provide a measure for tsetse prevalence , provide evidence for the impact of an intervention strategy or could be a sensitive tool to detect re-invasion of previously cleared areas or the efficacy of a barrier protecting a cleared area . Here , rTsal1 and total G . morsitans saliva seem to provide an indication of exposure to a broad range of Glossina species . Only for G . f . fuscipes , detection of induced IgGs is hampered which could relate to the genetic distance of this species as member of the Palpalis group ( although G . palpalis gambiensis was efficiently cross-detected ) . Nevertheless , serological responses against rTsal1 were detected in humans from Uganda principally exposed to G . fuscipes , which suggests that sensitivity and specificity parameters related to a single antigen may vary considerably depending on the host species . It also remains to be evaluated whether the intensity and persistence of the anti-Tsal1 IgG responses would not result in a saturation in populations persistently exposed to tsetse bites . This could possibly limit application of the rTsal1-based assay to examining subjects in tsetse-free areas that are at risk of invasion or to monitoring sentinel animals in order to evaluate tsetse presence in endemic regions . Exposure to the saliva of other hematophagous insects ( Tabanus and Stomoxys sp . ) that may be abundant in areas where tsetse flies are present , revealed that G . m . morsitans saliva did not yield false positive signals . Observations of Stomoxys and Tabanus exposed cattle revealed that also G . m . submorsitans saliva shares this feature of specificity , while Glossina tachinoides and G . palpalis gambiensis saliva yielded unspecific reactions with Tabanus exposed plasmas [20] . When using rTsal1 as an antigen , a weak unspecific signal was observed with the 1∶100 to 1∶800 diluted plasmas of mice exposed to Stomoxys calcitrans saliva which could be due to the presence of a homologue ( GenBank Accession N°: ACN69159 , [26] ) with relatively limited degree of identity to Tsal1 . Given the strong immunogenicity of the Tsal proteins and based on our ELISA analyses with the experimentally exposed pigs and mice we propose to use a 1∶1600 plasma dilution in favor of an increased specificity and an overall reduced risk of detecting cross-reactive antibodies . For the porcine ELISA , E . coli soluble extract was added to the sample diluent to increase the specificity similar to what was proposed for other pig serological tests [25] . Under these stringent experimental conditions , the recombinant Tsal1 was able to detect the high levels of exposure to tsetse fly bites and with a good correlation with the data obtained for total G . m . morsitans saliva as an antigen . Collectively and combined with our previous analyses using a large panel of East-African human plasmas [15] , this study indicates that a recombinant version of the G . m . morsitans Tsal1 fulfills the criteria of a candidate exposure biomarker for a broad range of tsetse fly species . We believe that the high sensitivity of the rTsal antigen and the broad species recognition could be an added value to the immunoassays that are tested in the framework of other studies based on tsetse salivary peptides . | Salivary proteins of hematophagous disease vectors represent potential biomarkers of exposure and could be used in serological assays that are complementary to entomological surveys . We illustrate that a recombinant version of the highly immunogenic Tsal1 protein of the savannah tsetse fly ( Glossina morsitans morsitans ) is a sensitive immunological probe to detect contact with tsetse flies . Experimental exposure of mice and pigs to different regimens of tsetse fly bites combined with serological testing revealed that rTsal1 is a sensitive indicator that can differentiate the various degrees of exposure of animals . Tsetse-induced antibodies persisted relatively long , and an efficient boosting of immunity was observed upon re-exposure . Recombinant Tsal1 is a promising candidate to detect contact with various tsetse species , which would enable screening of populations or herds for exposure to tsetse flies in various areas on the African continent . This exposure indicator could be a valuable tool to monitor the impact of vector control programs and to detect re-invasion of cleared areas by tsetse flies . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"veterinary",
"immunology",
"veterinary",
"diseases",
"zoonoses",
"medicine",
"and",
"health",
"sciences",
"epidemiology",
"african",
"trypanosomiasis",
"vector",
"biology",
"immunoassays",
"biology",
"and",
"life",
"sciences",
"immunologic",
"techniques",
"trypanosomiasis",
"disease",
"vectors",
"veterinary",
"science",
"veterinary",
"medicine",
"research",
"and",
"analysis",
"methods"
] | 2014 | Serological Responses and Biomarker Evaluation in Mice and Pigs Exposed to Tsetse Fly Bites |
The functions of many schistosome gene products remain to be characterized . A major step towards elucidating function of these genes would be in defining their sites of expression . This goal is rendered difficult to achieve by the generally small size of the parasites and the lack of a body cavity , which precludes analysis of transcriptional profiles of the tissues in isolation . Here , we describe a combined laser microdissection microscopy ( LMM ) and microarray analysis approach to expedite tissue specific profiling and gene atlasing for tissues of adult female Schistosoma japonicum . This approach helps to solve the gene characterization “bottle-neck” brought about by acoelomy and the size of these parasites . Complementary RNA obtained after isolation from gastrodermis ( parasite gut mucosa ) , vitelline glands and ovary by LMM were subjected to microarray analyses , resulting in identification of 147 genes upregulated in the gastrodermis , 4 , 149 genes in the ovary and 2 , 553 in the vitellaria . This work will help to shed light on the molecular pathobiology of this debilitating human parasite and aid in the discovery of new targets for the development of anti-schistosome vaccines and drugs .
Members of the genus Schistosoma are parasitic blood flukes responsible for the serious but neglected human disease of schistosomiasis [1] , [2] . In common with other platyhelminths , schistosomes exhibit acoelomy , the body plan characteristic of basal bilaterians whereby tissues are bound together by cells and matrices of the parenchyma in the absence of a body cavity . This body organization , together with the generally small size of adults and developing stages , has been a major hindrance for functional analyses of individual schistosome tissues and cells , because it has been impossible to isolate them . These problems are exacerbated by poor knowledge and limited annotations of many schistosome genes and the absence of basic knowledge of where , and when , in development the molecules are expressed . Localization methods incorporating immunocytochemistry and in situ hybridization have been at the vanguard of functional studies of schistosome proteins [3] , but the prospect of obtaining robust , informative localization data of multiple genes expressed throughout the complex schistosome life cycle remains a daunting challenge . Concerted international efforts have been directed at defining functional relevance of the predicted 14–16 , 000 schistosome genes to identify potential targets for drug and vaccine therapies [4] , [5] . Release of extensive schistosome ESTs ( Expressed Sequence Tags ) datasets and the anticipated publication of complete genomes for Schistosoma mansoni and S . japonicum [5] , [6] , [7] have provided new stimulus to achieving these goals . These datasets have enabled development of platforms for transcriptome and proteome analyses to explore gender , developmental and strain differences in schistosomes [8] , [9] , [10] , [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] . Here , we report on tissue-specific gene expression analysis of adult female Schistosoma japonicum , as a means to expedite functional characterization of schistosome gene products . Our approach incorporates methods of laser microdissection microscopy ( LMM ) to generate tissue-specific transcriptional extracts for subsequent microarray analysis . This work follows hypotheses [19] , [20] that LMM would prove an excellent means to expedite transcriptional typing of many schistosome tissues despite the acoelomate body plan of these parasites .
The use of mice in this study was approved under Project P288 by the Animal Ethics Committee of the Queensland Institute of Medical Research . Schistosoma japonicum-infected Oncomelania hupensis hupensis snails , collected from Anhui Province , China , were provided by the National Institute of Parasitic Diseases-CDC , Shanghai . Adult worm pairs were perfused 6 weeks post-challenge from infected ARC Swiss mice . Two batches of approximately 25 live female parasites were flat embedded in Tissue-Tek Optimal Cutting Temperature compound ( OCT ) ( ProSciTech , Australia ) and snap-frozen on dry ice . The sample blocks were stored at −80°C , prior to sectioning with sterile blades in a cryostat . Sections were cut at 7 µm and mounted onto a sterilized polyethylene-naphthalene membrane on a microscope slide ( P . A . L . M . Microlaser Technologies , Germany ) . The slides were then stored at −80°C . For transmission electron microscopy , female parasites were fixed in 3% glutaraldehyde in 0 . 1 M phosphate buffer at pH 7 . 4 for 2 h , post-fixed in potassium ferricyanide-reduced osmium tetroxide , followed by 5% aqueous uranyl acetate , dehydrated in acetone and embedded in Epon resin . Ultrathin sections were viewed on a JEM 1011transmission electron microscope operated at 80 kV . Thawed cryo-sections were fixed immediately in 100% methanol for 30 seconds , stained with 1% Toluidine blue ( CHROMA , Germany ) for 10 seconds , washed in diethylpyrocarbonate ( DEPC ) -treated water , 2×10 seconds , and allowed to dry for 20–30 min before microdissection . A PALM microbeam laser catapult microscope ( P . A . L . M . Microlaser Technologies ) was used to microdissect the gastrodermis from posterior regions of female worms , and the ovary and vitelline tissues from the stained frozen sections . An area of approximately 4 million squared µm ( approximately 20×106 µm3 of tissue ) was collected separately from each of the tissues onto 500 µl opaque adhesive caps ( P . A . L . M . Microlaser Technologies ) . The areas amount to the collection of many 1000's of microdissected elements for each tissue . For control tissue , we used 12 S . japonicum females that were snap-frozen in OCT and sectioned by cryostat . Control sections were collected onto 6 sterile glass slides . Entire sections were then scraped by sterile scalpel blades from the slides into RNA extraction buffer ( below ) for analysis . The control samples therefore represent the transcriptional repertoire of entire females . Total RNA was isolated from the control and LMM samples using RNAqueous-Micro kits ( Ambion ) kit using the manufacturer's instructions and quantified using a Nano-Drop ND-1000 spectrophotometer ( Thermo Scientific , USA ) . The quality of total RNA was assessed using a Bioanalyser RNA Pico Lab Chip ( Agilent ) prior to storage at −80°C . Full details of the design and construction of the schistosome microarray used have been reported [11] . In brief , the array was constructed from information based on the transcriptomes of adult S . japonicum and S . mansoni . The microarray consists of 19 , 222 target contiguous sequences ( contigs ) printed twice from two independent probe designs , and includes 12 , 166 probes derived from S . mansoni , and 7 , 056 probes derived from S . japonicum . An overview of the design and composition of the microarray is present in Table S1 . A 300 ng aliquot of total RNA from each sample was converted into complementary RNA was synthesized and labeled with the fluorophore Cyanine 3-CTP ( CY3c ) and hybridized according to the manufacturer's instructions ( Agilent Technologies -One-Color Microarray-Based Gene Expression Analysis ) . Microarray hybridisations were performed in duplicate for all samples . The yield , concentration , amplification efficiency and abundance of CY3c were measured at A260 and A550 by spectrophotometry . Hybridized slides were scanned using an Agilent Microarray Scanner ( B version ) as tiff files and processed with the Feature Extraction 9 . 5 . 3 . 1 Image Analysis program ( Agilent ) to produce standardised data for statistical analysis . All microarray slides were assessed for background evenness by viewing the tiff image by Feature Extraction . Feature extracted data was analysed using GENESPRING ( version 7 . 3 . 1; Agilent Technologies/Silicon Genetics , Redwood City , CA ) . Microarray data were normalised using a normalisation scenario for “Agilent FE one-color” which including “Data Transformation: Set measurements less than 5 . 0 to 5 . 0” , “Per Chip: Normalize to 50th percentile” and “Per Gene: Normalize to median” . Data sets were further analysed using published procedures based on one-colour experiments [21] . The gProcessedSignal values determined in GENESPRING using Agilent's Feature Extraction software including aspects of signal/noise ratio , spot morphology and homogeneity . Thus , gProcessedSignal represents signal after localised background subtraction and includes corrections for surface trends . Features were deemed Absent when the processed signal intensity was less than two fold the value of the processed signal error value . Features were deemed Marginal when the measured intensity was at a saturated value or if there was a substantial amount of variation in the signal intensity within the pixels of a particular feature . Features that were not Absent or Marginal were deemed Present . Data points were included only if Present or Present , Absent and probes or contigs retained if all data points were Present or Present , Absent . Microarray data have been submitted to the Gene Expression Omnibus public database , under accession numbers GPL7160 and GSE12706 . Batch BlastX ( 6 frame translation protein homology ) was performed at http://www . blast2go . de on all contigs . This presented a further overview of the gene ontologies that are modulated between tissue types in adult female S . japonicum ( Figure S1 and Table S2 ) . This information was used to supplement previously published GOs based on nucleotide sequence [11] . To gain a more complete overview of the GO categories that are modulated during the S . japonicum lifecycle we used the software ErmineJ to produce extended list of GOs associated with each of the microdissected tissue types [22] . A total of 9 gene sequences indentified as differentially expressed among the three S . japonicum tissues and whole worm control tissue were chosen for validation of microarray data using real time PCR as described [12] . The template for real time PCR was that obtained by microdissection . Forward and reverse primers ( Sigma-Aldrich , Australia ) were designed from the 10 contigs ( Table S3 ) . All total RNA samples were DNase treated ( Promega , Australia ) prior to synthesis of cDNA using a QuantiTect Whole Transcriptome Kit ( QIAGEN , Australia ) . All cDNA samples were diluted to a concentration of 5 ng/µl . Real time PCR was performed in a Gene Disc 100 ring ( Corbett Research , Australia ) . A sequence from the NADH-ubiquinone reductase gene of S . japonicum was used for normalisation of data from all experiments . Each experiment was performed in duplicate , and the confidence threshold ( CT ) of the second set was normalised to the first set before evaluation . This was done by importing the standard curve of the first set to that of the second using Rotor Gene 6 software [12] . Microarray and real time PCR datasets were tested following Morley and colleagues [23] . Data were analysed using Graphpad Prism Version 5 . Data from microarray and real time PCR populations were examined to ascertain if they fit normal distributions , using the D'Agostino and Pearson omnibus normality test and the Shapiro-Wilk normality test . Because both sets of data were not normally distributed , a Spearman correlation ( Rho ) was employed to test for correlation . The statistical analyses used an alpha value of 0 . 05 .
We targeted three female tissues , namely , gastrodermis ( absorptive gut lining ) of the posterior halves of the worms , ovary , and the vitelline glands ( = vitellaria , accessory glands of the female system that produce precursors for eggshell synthesis ) ( Figures 1 and 2 ) . We chose these three tissues due to their relative abundance , clearly delimited structure and the important biological roles in schistosome development and reproductive biology . In view of the closely knit organization of schistosome tissues , it was important to know whether the three tissues under investigation represented homogenous cell populations . Ultrastructural assessment indicated that the ovary and gastrodermis were homogenous ( Figure 3 ) . We had previously shown through ultrastructural studies incorporating a stereological analysis of the relative volumes of tissues in vitellogenic regions that although some parenchymal tissues intrude into the vitelline regions , vitellogenic regions are dominated by vitelline cells ( vitellocytes ) [24] which are highly synthetic cells . Thus , all tissue extracts represent homogenous or near homogenous samples . For microarray analysis , unfixed frozen females were sectioned by cryostat onto membrane-coated slides , stained with toluidine blue and microdissected using a PALM laser catapult microscope ( Figure 1 ) . Total RNA integrity from microdissected samples was assessed ( Figure 1 ) and shown to be of high fidelity . A distinct 28S band is never visible in total RNA fractions of schistosomes [25] . RNA was further processed for one-colour fluorophore-labelled cRNA synthesis and hybridization to a microarray representing the near complete transcriptome of adult schistosomes [11] . Of 38 , 444 probes ( representing 19 , 222 contigs ) on the chip , 8 , 454 ( 5 , 242 contigs ) were retained after filtering ( Table S4 ) . Principal component analysis ( PCA ) is a multi-dimension reduction method that allows the visual presentation of a complex data set , so that distances between plotted points represents the relative similarity of each datasets . Usually plotted in an X , Y , Z formation , each axis represents a distinct subset of data points , or in the current application , gene lists . Gene expression profiles of the three microdissected tissues and the control sample were analysed by PCA ( Figure S2 ) . The point of the control tissue was more similar to those of the gastrodermis and vitellaria , compared with the ovary . This observation is not surprising , for the former tissues are voluminous in female parasites and likely account for much of the female transcriptome . Complete lists of genes enriched for each tissue sample after normalization , together with lists of selected genes of interest enriched for each tissue are presented ( Tables 1–3 , Table S4 , and Figure 2 ) . Major gene ontologies ( GOs ) of differentially expressed genes for the three tissues are also shown ( Figure S1 and Table S2 ) . Abundant transcripts enriched for each tissue encoded protein sequences for which there was little or no annotation or sequence identity . Nine transcripts that were enriched in one of the 3 tissues were selected for validation of expression level by real time PCR using cDNA templates from the microdissected and control samples ( Figure 4 ) . Expression levels observed by real time PCR agreed with those by microarray for these genes . The microarray and real time PCR data sets of the 9 genes showed a significant correlation of 0 . 6791 ( Spearman's Rho , p<0 . 0001 , n = 27 ) . After filtering the microarray data and normalizing signal relative to female germinal tissues , we identified 214 probes representing 147 genes enriched for the gastrodermis ( Figure 2 , Table 1 , and Table S4 ) . Comparable datasets in Table S4 compare gene expression of the gastrodermis relative to either ovary or vitellaria . These three datasets show strong congruence , although with some variation in relative enrichment of some sequences . Thus , a ferritin isoform is enriched in the gastrodermis relative to the ovary , but not relative to the vitellaria ( Contig7767 , Table S4 ) . The enriched genes of the gastrodermis relative to female germinal tissues included proteases of the haemoglobinolytic cascade , membrane-associated molecular transporters , actin and associated molecular motors . A highly enriched gene of the gastrodermis , represented by Contig5007 , is a hitherto uncharacterized gene with uncertain sequence identity , but which contains motifs with similarity to the meprin family of metalloproteinases , and an erythrocyte-binding protein of malaria parasites . This molecule potentially represents a novel class of proteinases involved in haemoglobinolysis in these vascular parasites [26] . Surprisingly , cathepsin D , an early member of the haemoglobinolysis cascade [27] , was not enriched for the gastrodermis . Given its upstream role in this multi-enzyme network , cathepsin D is probably expressed in anterior zones of the gut , either in the oesophageal gland , or in anterior zones of the gastrodermis . Our study focused on microdissection of the posterior regions of the gastrodermis . Regional specialization of the apparently simple gastrodermis of other platyhelminths has been postulated [28] . It may be that the schistosome gut displays a similar planar polarity , evidenced by distinct secretory product in different zones along the length of the parasite [26] . The hypothesis is further substantiated by observations that the gastrodermal regions analysed here were enriched for numerous sequences encoding dipeptidases and carboxypeptidases ( Table S4 ) , peptidases more likely to be associated with terminal parts of the haemoglobinolytic cascade . Additionally , transcripts encoding proteins previously localized to the outer tegumental surface of the parasite ( tetraspanins , annexin and alkaline phosphatase ) were enriched for the gastrodermis relative to other tissues . Although these molecules have been previously recognised as tegumentary components , their occurrence in the syncytial gastrodermis is not surprising . Transcripts for divalent metal transporters , particularly a member of the Zinc regulated transporter/iron regulated transporter family ( ZIP ) family were enriched for the gastrodermis . Schistosomes have high dietary requirements for iron [29] , [30] and other divalent metals . While a surface mediated pathway for iron uptake by schistosomes has been postulated [29] , the presence of metallo-transporters in the parasite gut indicates that this tissue may also scavenge the trace metals [29] . Other transcripts enriched for the gastrodermis ( relative to ovary and vitelline tissues ) represent genes encoding lysosomal proteins , namely , cystinosin , lysosomal acid membrane glycoprotein ( Lamp1/CD68 ) , lysosomal alpha mannosidase and acid phosphatases ( Figure 2 , Table 1 , and Table S4 ) , although lysosomes are not abundant cytoplasmic features of the gastrodermis . A consistent feature of the syncytium , however , is the presence of apical epicellular vacuoles ( Figure 3 ) , which enclose parts of ingested host blood , and are lined by villus-like lamellae . The vacuoles have many features of lysosomes , namely a low pH and presence of proteolytic enzymes [26] and are a possible cellular location for the lysosome molecules of the gastrodermis . We identified 6 , 645 probes representing 4 , 149 upregulated genes ( Figure 2 , Table 2 , and Table S4 ) for the ovary compared with the gastrodermis . Similarly , we identified 3 , 832 probes representing 2 , 553 upregulated genes ( Figure 2 , Table 3 , and Table S4 ) for vitellaria compared with the gastrodermis . Oocytes and vitellocytes , in platyhelminth evolution and ontogeny , are believed to be derived from common progenitor cells [31] . We decided , therefore , to determine whether the tissues have common expression identity that may reflect the common origin of the two tissues . Analysis of expression by Venn diagram indicated substantial overlap in expression between the two germinal tissues , but not with the gastrodermis ( Figure 2B ) . Genes enriched for both cell types included egg-specific genes including major egg antigens and egg protein cp422 . The former gene is also abundant in mature eggs of schistosomes [32] . Other genes enriched for the female reproductive tract included those encoding molecules for TGF-β and tyrosine kinase signalling pathways [33] , [34] , different innexins ( gap junction proteins of invertebrates ) , and a diversity of genes encoding molecules associated with DNA processing , replication , and transcription . With the exception of the egg-specific antigens , the upregulated genes common to the two tissues are involved in cell proliferation and intercellular signalling . Genes enriched for the ovary ( relative to the gastrodermis ) included a number encoding proteins associated with cytokinesis , fertilization and coated pit-mediated endocytosis ( Table 2 ) . Oocytes express genes with identity to polycomb , enhancers of polycomb , and Peter pan homologues of vertebrates and ecdysozoans [35] ( Table 2 ) . Polycomb genes , not previously recognized for platyhelminths , repress Hox expression in embryogenesis leading to cellular and zonal differentiation in embryos . Discovery of genes involved in embryonic differentiation will provide new insight into developmental cascades in the complex multi-generational schistosome life-cycle , leading in turn to a better understanding of differentiation of the intraovular embryo , the stage primarily responsible for pathogenesis in schistosomiasis . Expression analysis of vitellaria ( Table 3 , Figure 2 , and Table S4 ) revealed enriched genes ( relative to the gastrodermis ) associated with egg-shell synthesis , as well as a range of membrane transporters with affinity for amino acids , metallo-ions and nucleotides . Eggshell precursors , egg-specific proteins and tyrosinases were upregulated as expected for this tissue that provides precursors for choriogenesis [30] , [36] . Numerous membrane-spanning transporters and genes encoding proteins for exocytosis were also enriched as were those associated with lipid metabolism . Some transcripts , annotated as containing signal peptides , did not contain abundant tyrosine residues , a prerequisite for eggshell precursors [37] , possibly indicating that these molecules function in aspects other than shell formation . Given the essential role of vitellocytes in egg development and embryogenesis , functional characterization of these putative secreted proteins may enhance our understanding of the complexity of egg-shell synthesis and may help resolve long-standing questions about yolk function of vitellocytes [30] . The integration of microarray analysis of LMM-dissected tissues has provided the means to establish a gene expression atlasing strategy for S . japonicum , alleviating the technology hurdles imposed by the acoelomate nature of this platyhelminth and expediting localization of multiple genes . Tissue-specific expression profiling has been performed previously for cavitate invertebrates , incorporating LMM or gross dissection methods [38] , [39] , but this approach demonstrates the feasibility for gene mapping in a platyhelminth , thus serving as an exemplar for similar studies of other basal bilaterians and small organisms . The localization data provided here serves as a novel resource to advance functional studies of many unannotated S . japonicum genes , thereby providing a valuable molecular platform to shed light on the complex physiology and biochemistry of schistosomes , the pathogenesis of schistosomiasis , and to develop new treatments and effective interventions for its control . | Schistosomes are parasitic worms responsible for important human diseases in tropical and developing nations . There is urgent need to develop new drugs and vaccines to augment current treatments for this disease . In recent years , concerted efforts by many laboratories have led to extensive genetic sequencing of the parasites , and the publication of genome sequence for two agents of schistosomiasis appears imminent . This genetic information has revealed many molecules expressed by the schistosome parasites for which no functional information is available . This lack of information extends to ignorance of where in the complex multicellular schistosome parasites the genes are expressed . We integrated two molecular and cellular techniques to address these knowledge gaps . We used laser microdissection microscopy to dissect small but highly important tissues involved in nutrition and reproduction from sections of female Schistosoma japonicum . From these dissected tissues we then used a broad molecular biology method to identify the multiple genes active in these tissues . Our approach has allowed us to formulate the basis of a “gene atlas” for schistosome parasites , defining the expression repertoire of specific tissues . The better understanding of the roles of tissues in parasite biology , especially in development , reproduction and interactions with its human hosts , should promote future investigations into pathogenesis and control of these significant parasites . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results/Discussion"
] | [
"molecular",
"biology/bioinformatics",
"infectious",
"diseases/helminth",
"infections",
"molecular",
"biology/molecular",
"evolution",
"infectious",
"diseases/neglected",
"tropical",
"diseases"
] | 2009 | Tissue Specific Profiling of Females of Schistosoma japonicum by Integrated Laser Microdissection Microscopy and Microarray Analysis |
Leishmania parasites replicate within the phagolysosome compartment of mammalian macrophages . Although Leishmania depend on sugars as a major carbon source during infections , the nutrient composition of the phagolysosome remains poorly described . To determine the origin of the sugar carbon source in macrophage phagolysosomes , we have generated a N-acetylglucosamine acetyltransferase ( GNAT ) deficient Leishmania major mutant ( ∆gnat ) that is auxotrophic for the amino sugar , N-acetylglucosamine ( GlcNAc ) . This mutant was unable to grow or survive in ex vivo infected macrophages even when macrophages were cultivated in presence of exogenous GlcNAc . In contrast , the L . major ∆gnat mutant induced normal skin lesions in mice , suggesting that these parasites have access to GlcNAc in tissue macrophages . Intracellular growth of the mutant in ex vivo infected macrophages was restored by supplementation of the macrophage medium with hyaluronan , a GlcNAc-rich extracellular matrix glycosaminoglycan . Hyaluronan is present and constitutively turned-over in Leishmania-induced skin lesions and is efficiently internalized into Leishmania containing phagolysosomes . These findings suggest that the constitutive internalization and degradation of host glycosaminoglycans by macrophages provides Leishmania with essential carbon sources , creating a uniquely favorable niche for these parasites .
Protozoan parasites , belonging to the genus Leishmania , cause a number of important diseases in humans , that affect over 12 million people worldwide with more than 2 million new infections each year [1] . Clinical disease ranges from self-healing cutaneous lesions to visceral leishmaniasis , which is invariably fatal if left untreated . There is currently no effective vaccine against Leishmania and front-line drug treatments are limited by high toxicity , expense , requirement for hospitalization and the emergence of drug resistance [2] . Leishmania are transmitted by a sandfly vector that injects infective metacyclic promastigotes into the skin of the mammalian host during a bloodmeal . Promastigotes are phagocytosed by macrophages that are recruited to the site of the sandfly bite , either directly or after passage through neutrophils [3 , 4] . Parasites internalized by macrophages are delivered to the mature phagolysosome compartment where they differentiate to the non-motile amastigote stage . While the biogenesis of the Leishmania-occupied phagolysosome has been intensively studied [5] , relatively little is known about the nutrient environment within this compartment or the specific carbon sources utilized by these parasites . As phagocytosis and subsequent lysosomal degradation constitutes one of the major functions of macrophages , it is likely that the nutrient composition of the phagolysosome will vary depending on the physiological state and activity of host macrophage [6] . During early stages of infection , Leishmania likely reside in macrophages that are involved in wound repair and tissue remodeling processes [7] , while at later stages of infection they appear to replicate in alternatively activated macrophages that display distinct microbicidal responses and metabolism [8 , 9] . Leishmania also reside within other phagocytic host cells that disseminate to lymph nodes and other organs [5] . However , little is known about the metabolic environment that sustains parasite life under these conditions . Previous studies have shown that Leishmania amastigotes are dependent on the up-take of sugars for growth and virulence in the mammalian host [10] . Specifically , targeted deletion of three high affinity hexose transporters in L . mexicana severely reduces intracellular growth of amastigote in macrophages [11 , 12] . Similarly , disruption of a L . major gene encoding the enzyme , glucosamine-6-phosphate deaminase ( GND ) , required for the catabolism of amino sugars , also results in strong attenuation of amastigote growth in both macrophages and in susceptible mice [13] . The dependence of L . major amastigotes on GND indicated that these stages have access to amino sugars , such as glucosamine ( GlcN ) or N-acetylglucosamine ( GlcNAc ) within the phagolysosome compartment . This is consistent with the finding that a L . major mutant , auxotrophic for all amino sugars , was able to infect mice and establish normal lesions [14] . Amino sugars are present as free sugars in the blood , and are also major components of a number of mammalian glycoconjugates , including glycoproteins ( N- and O-glycans/glycosylphatidylinositol anchors ) , proteoglycans and glycolipids [15] . However , the extent to which Leishmania have access to or can utilize host glycoconjugates as a carbon source has not been investigated . To further investigate the capacity of Leishmania to salvage specific amino sugars , we have generated a L . major mutant that is a strict auxotroph for GlcNAc . This mutant was capable of establishing large skin lesions in susceptible mice , but failed to survive in cultured macrophages . Analysis of the growth phenotype of this mutant in ex vivo infected macrophages showed that intracellular growth was not dependent on the uptake of free amino sugars , but rather was rescued by supplementation of infected macrophages with high molecular weight hyaluronan , an abundant GlcNAc-rich polysaccharide component of the extracellular matrix . We propose that uptake and degradation of hyaluronan provides Leishmania amastigotes with essential carbon sources , suggesting that the parasite exploits a major function of macrophages in extracellular matrix turnover and remodeling in the skin and other tissues .
L . major promastigotes can synthesize amino sugar phosphates de novo via the hexosamine biosynthesis pathway that includes the enzymes , glutamine:fructose-6-phosphate amidotransferase ( GFAT ) and GNAT ( Fig 1 ) . Targeted deletion of GFAT , the first enzyme in this pathway , results in amino sugar auxotrophy that can be bypassed by supplementation of the medium with either GlcN or GlcNAc ( Fig 1 ) [14] . In contrast , targeted deletion of GNAT , the second enzyme in this pathway , would be expected to lead to amino sugar auxotrophy that could only be bypassed by exogenous GlcNAc , but not GlcN . The L . major genome contains a gene encoding a putative GNAT ( LmjF28 . 3005 ) that shares 48% similarity and 28% identity at the amino acid level to the yeast GNAT ( ScGNA1p ) ( S1 Fig ) . Importantly , amino acids involved in substrate and cofactor binding ( Glu98 and Asp99 , located in a large hydrophobic cleft ) and enzyme catalysis ( Tyr143 ) in the yeast enzyme are conserved in the L . major GNAT ( S1 Fig ) [16] . Closely related GNAT homologues are also present in the genomes of the trypanosomatid parasite T . cruzi ( S1 Fig ) and have recently been shown to be essential for blood stage T . brucei [17] . The targeted deletion of the L . major GNAT was achieved by sequential replacement of the two chromosomal alleles with SAT and BLEO resistance cassettes by double homologous recombination ( S2A Fig ) . Clones lacking the GNAT gene were readily isolated when parasites were cultivated in rich medium containing GlcNAc and loss of the GNAT gene confirmed by PCR analysis ( S2B Fig ) . As expected , the growth of the ∆gnat mutant was dependent on the presence of exogenous GlcNAc ( Fig 2A ) and complete loss of viability was observed after 72 hours in the absence of GlcNAc ( S2C Fig ) . In contrast , GlcN , which is readily taken up by promastigotes and converted to GlcN6P ( the substrate for GNAT ) [14] was unable to rescue growth ( Fig 2A ) . Importantly , GlcNAc prototrophy in the mutant was restored by ectopic expression of GNAT from the pXG-PURO plasmid ( Fig 2B ) . To assess the minimum amount of GlcNAc required for growth , L . major ∆gnat promastigote growth was determined in media supplemented with decreasing concentration of GlcNAc ( 0–5 μg/ml ) over time . L . major ∆gnat parasites displayed normal growth kinetics when medium was supplemented with 5 μg/ml of GlcNAc , while no growth was observed when parasites were supplemented with 1 μg/ml GlcNAc ( Fig 2C ) . Addition of excess sugar to specific Leishmania sugar auxotrophs can lead to toxicity as a result of the hyper-accumulation of the cognate sugar phosphate and depletion of ATP [18] . However , addition of 50–500 μg/ml GlcNAc to the growth media had no detrimental impact on parasite growth ( Fig 2D ) , indicating that high GlcNAc levels are not toxic to L . major ∆gnat promastigotes . To confirm that deletion of L . major GNAT results in loss of GNAT activity , cell lysates of wild type , ∆gnat and complemented parasite lines were incubated with GlcN6P and the biosynthesis of GlcNAc6P assessed by direct infusion electro-spray ionization mass spectrometry . Synthesis of GlcNAc6P , with a concomitant decrease in GlcN6P levels , was detected in lysates of wild type and complemented parasite lines , but not in lysates generated from the ∆gnat mutant ( Fig 3A ) . The single L . major GNAT gene therefore appears to account for all of the GNAT activity . As GlcNAc is a core component of free GPI and GPI-anchor glycolipids , the synthesis of these glycoconjugates in the ∆gnat mutant should be dependent on the presence of exogenous GlcNAc . Indeed , when ∆gnat promastigotes were labeled with 3H-glucose in the absence of exogenous GlcNAc , the synthesis of free GPIs , but not phospholipids , was completely abrogated ( Fig 3B ) . The GlcNAc-starved ∆gnat mutant was also deficient in the expression of the major surface glycoconjugates , LPG and gp63 , which are both anchored to the plasma membrane via GPI glycolipids ( Fig 3C ) . As expected , the synthesis of the major intracellular reserve carbohydrate , mannogen , was not abrogated in ∆gnat promastigotes following removal of GlcNAc ( Fig 3D ) . Rather , increased mannogen levels were observed under these conditions ( Fig 3D ) , which may reflect the diversion of excess hexose phosphates into mannogen synthesis in the absence of parasite growth . GlcNAc is also a core component of N-linked glycans that are assembled in the ER as lipid-linked oligosaccharides ( LLO ) . The cellular levels of LLO precursor pools can be assessed by incubating Leishmania lysates with GDP-3H-Man , which results in the rapid labeling of preexisting LLO precursors . As shown in Fig 3E , removal of GlcNAc from the culture medium had no effect on the LLO pool size in wild type parasites , but resulted in the rapid depletion of LLO precursors in the ∆gnat mutant . Collectively , these results demonstrate that GNAT is required for the synthesis of essential glycoconjugates in the absence of exogenous GlcNAc . The virulence of ∆gnat parasites was assessed in the highly susceptible BALB/c mouse model of infection . Infections were initiated with stationary phase wild type and ∆gnat promastigotes cultivated in the presence of GlcNAc , which contained similar levels of metacyclic parasites ( Fig 4A ) . Wild type parasites induced lesions within 3–4 weeks , which increased in size over time ( Fig 4B ) . ∆gnat parasites also induced large lesions , comparable in size and severity to wild type parasites . However , lesion development was reproducibly delayed by 4–5 weeks ( Fig 4B ) . This delay was also observed with ∆gnat promastigotes differentiated from mature lesions , suggesting that it is not a consequence of culture-induced loss of virulence . Unexpectedly , while complementation of the ∆gnat mutant reversed the mutant phenotype in vitro , it resulted in complete loss of virulence in mice ( Fig 4B ) , raising the possibility that expression of GNAT under non-native conditions is deleterious in vivo . In contrast to promastigotes stages , isolated ∆gnat amastigotes re-established lesions in mice without a lag phase and comparable kinetics to wild type parasites ( Fig 4C ) . The ∆gnat parasites were able to salvage sufficient levels of GlcNAc from the host , as lesion-derived amastigotes expressed free GPIs ( GIPL-1 , 2 and 3 ) , although at reduced levels compared to wild type amastigotes ( S3A Fig ) . GPI biosynthesis in these amastigotes was not due to restoration of GlcNAc synthesis as isolated ∆gnat parasites lacked the GNAT gene , as determined by PCR ( S3B Fig ) and were unable to grow as promastigotes in the absence of exogenous GlcNAc ( S3C Fig ) . These results suggest that while phagolysosomes harboring promastigotes may have limiting levels of GlcNAc , they contain sufficient levels of this amino sugar to promote ∆gnat amastigote differentiation and growth and skin lesion formation . To further investigate potential sources of GlcNAc utilized by intracellular amastigotes , infection experiments were undertaken in BALB/c bone marrow-derived macrophages ( BMDMs ) . ∆gnat promastigotes were rapidly cleared by BMDMs within four days post infection . In contrast , wild type parasite levels remained constant over the same period ( Fig 4D ) . Similar to promastigotes , lesion derived ∆gnat amastigotes were unable to survive in cultured macrophages ( Fig 4E ) , suggesting that ex vivo macrophages , but not macrophages in skin lesions , fail to provide sufficient levels of GlcNAc to support intracellular ∆gnat growth . To investigate whether intracellular survival of ∆gnat parasites could be rescued by exogenous GlcNAc , the medium of infected macrophages was supplemented with 50 μg/ml or 500 μg/ml GlcNAc . Addition of physiologically relevant concentrations of GlcNAc ( ~40 μg/ml in serum [15] ) did not rescue growth of the ∆gnat mutant , while addition of a 10-fold higher concentration GlcNAc resulted in a modest increase ( 2-fold ) in intracellular growth ( Fig 5A ) . These results suggest that the fluid phase uptake of exogenous GlcNAc and/or transport of GlcNAc from the macrophage cytoplasm to the phagolysosome contribute minimally to the observed ∆gnat growth in skin lesions . The extracellular matrix of the dermis contains several glycosaminoglycans , including chondroitin sulfate , heparin and hyaluronan . Hyaluronan is the most abundant of these and the only glycosaminoglycan to contain unmodified GlcNAc within the repeat disaccharide units ( GlcAβ1-3GlcNAcβ1–4 ) n , of the polysaccharide backbone . As previous studies have shown that hyaluronan is constitutively turned-over by macrophages in the skin [19] , we investigated whether exogenous hyaluronan can rescue the intracellular growth phenotype of growth of ∆gnat parasites . Addition of high molecular weight hyaluronan to ∆gnat-infected macrophages greatly enhanced intracellular survival of ∆gnat over and above supplementation of infected macrophages with free GlcNAc ( Fig 5A ) . Addition of hyaluronan also resulted in a small but significant stimulation of intracellular growth of wild type parasites ( Fig 5A ) . In contrast , addition of chitin , a polysaccharide containing exclusively GlcNAc , did not increase intracellular ∆gnat survival ( Fig 5A ) . To investigate whether the different growth stimulatory properties of hyaluronan and chitin reflected differences in the capacity of the parasites to degrade these polymers , wild type and ∆gnat promastigotes were cultured in medium supplemented with GlcNAc , chitin or hyaluronan . Addition of these supplements had no growth inhibitory effect on the growth of wild type parasites ( Fig 5B ) . However , chitin but not hyaluronan had a strong stimulatory effect on ∆gnat growth in vitro , in contrast to the situation in ex vivo infected macrophages ( Fig 5C ) . The capacity of L . major parasites to utilize chitin is consistent with previous reports showing that promastigotes secrete a soluble chitinase [20 , 21] . Together , these results suggest that Leishmania are unable to degrade high molecular weight hyaluronan directly , but salvage free GlcNAc generated by phagolysosome hyaluronidases and other endo/exoglycosidases . Hyaluronan is present in the extracellular matrix as high molecular weight polymers ( 103−104 kDa ) which are largely anti-inflammatory . However , lower molecular weight polymers ( less than 50 kDa ) , which can be generated by extracellular hyaluronidases , induce expression of inflammatory cytokines in macrophages [22] . To exclude the possibility that intracellular ∆gnat rescue by hyaluronan was due to alterations in macrophage activation state , ∆gnat-infected macrophages were incubated with a range of hyaluronan polymers with defined molecular weights from 33 to 1500 kDa . Low molecular weight hyaluronan fractions ( modal molecular weight of 33 kDa ) failed to rescue intracellular ∆gnat parasites , while high molecular weight hyaluronan fractions ( from 158 to 1500 kDa ) were highly growth stimulatory ( Fig 6B ) . Growth of wild type parasite was also stimulated by high molecular weight hyaluronan polymers ( >500 kDa ) , although this result did not reach statistical significance ( Fig 6A ) . To further exclude the possibility that the growth promoting properties of hyaluronan was due to an indirect effect on macrophage activation , we tested the L . major ∆gnd mutant which lacks a key enzyme in GlcNAc catabolism [13] . As shown previously , this mutant is unable to replicate in macrophages , providing direct evidence that intracellular amastigote stages are dependent on GlcN or GlcNAc as a carbon source [13] . Strikingly , and in contrast to the ∆gnat mutant , intracellular growth of the ∆gnd mutant was not rescued by supplementation of the medium of infected macrophages with high molecular weight hyaluronan ( Fig 6C ) . These data suggest that hyaluronan internalization and degradation promotes intracellular parasite growth by providing an essential carbon source . The mammalian skin contains 50% of total body hyaluronan , which is turned over via endocytosis and degradation in lysosomes within 1–2 days [19 , 23] . To confirm and directly measure hyaluronan turnover in Leishmania skin lesions , we measured hyaluronan dynamics in animal tissues using 2H2O labeling [24] . Infected BALB/c mice were labeled with 2H2O in their drinking water over several days and the rate of hyaluronan turnover quantitated by measurement of the level of deuterium enrichment in the diagnostic tetrasaccharide repeat unit ( [GlcAβ1-3GlcNAcβ1–4]2 ) , after extraction , hyaluronidase depolymerization and analysis of the released oligosaccharide fragments by liquid chromatography-mass spectrometry ( LC-MS ) . Deuterium enrichment in the hyaluronan repeat unit increased rapidly during the first five days of labeling , reaching close to maximum labeling observed after four weeks labeling ( S4 Fig ) . These results suggest that hyaluronan is constitutively and rapidly turned over in Leishmania lesions with a half-life of <2 days ( S4 Fig ) , consistent with previous reports using healthy skin [23] . To confirm that hyaluronan is indeed internalized into Leishmania-containing vacuoles , infected macrophages were incubated with fluorescein-labeled hyaluronan ( HA::FITC ) . When macrophages were infected with L . major promastigotes expressing mCherry red fluorescent protein , green fluorescence was observed around the periphery of intracellular amastigotes that induce individual , tight fitting vacuoles ( S5A Fig ) . Green fluorescence was readily detected in vacuoles in cases where L . major induced more spacious compartments ( S5B Fig ) . Similar experiments were performed with macrophages infected with L . mexicana promastigotes that induce large communal phagolysosomes . HA::FITC accumulated in lysosomal compartments , as evidenced by co-localization with lysotracker , and the L . mexicana containing phagolysosome ( Fig 7 ) , demonstrating efficient endocytosis and lysosomal targeting of these polysaccharides in infected macrophages .
Leishmania parasites are dependent on the uptake of sugars for intracellular growth in the mammalian host [12 , 25] . However , the source of the sugars accessed by intracellular amastigotes has not been defined . We have previously shown that the L . major mutant lacking GND is unable to catabolize amino sugars as major carbon source and is heavily attenuated in mice infections [13] . In this study we have generated a new L . major sugar auxotroph that is specifically dependent on salvage of GlcNAc for growth and viability . Using this mutant as a probe for intracellular GlcNAc levels , we show that the levels of this sugar are limiting for parasite growth during early stages of infection , but not for amastigote growth in lesion macrophages . In ex vivo infected macrophages , intracellular ∆gnat amastigotes growth is efficiently restored by supplementation of infected macrophages with the GlcNAc-rich glycosaminoglycan , hyaluronan , but not by supplementation of the medium with physiological concentrations of GlcNAc . Importantly , we demonstrate that Leishmania induced skin lesions contain high levels of hyaluronan and that this glycosaminoglycan is constitutively turned over in situ . We therefore propose that the constitutive uptake and degradation of hyaluronan by macrophages may partly underlie the tropism for and capacity of Leishmania to proliferate within these host cells . We have previously shown that targeted deletion of L . major GFAT , the first enzyme in the hexosamine-phosphate biosynthetic pathway , had no measurable effect on amastigote growth and lesion development in mice [14] . The growth phenotype of ∆gfat in culture could be by-passed by the provision of either GlcN or GlcNAc , suggesting that either one or both sugars are present within the phagolysosome compartment of macrophages at sufficient levels to sustain amastigote growth at all stages of infection [14] . To further define the amino sugar composition of macrophage phagolysosomes and potential sources of these sugars , we generated the L . major ∆gnat mutant , which exhibits a restricted auxotrophy for GlcNAc alone . The L . major ∆gnat promastigotes showed a reproducible delay in skin lesion development , which was also observed with ∆gnat amastigotes at a low inoculum ( S6 Fig ) . Nevertheless , ∆gnat promastigotes and amastigotes induced skin lesions comparable to those induced by wild type parasites . These findings suggest that GlcNAc levels in infected macrophages are limiting during early stages of infection , but are sufficient to sustain normal parasite growth in mature lesions . It is possible that intracellular levels of GlcNAc in infected macrophages increase as lesions develop , possibly as a result of increased turnover of proteoglycans . Alternatively , lesion amastigotes may have a lower requirement for GlcNAc and/or other sugars than amastigotes during early stages of infection . In this respect , we have recently shown that Leishmania amastigotes switch to a metabolically quiescent state in lesions , which is characterized by markedly reduced rates of hexose uptake [24] . Taken together , the virulence phenotypes of different Leishmania hexose auxotrophs demonstrate that the mature phagolysosomes of macrophages contain sufficient levels of the amino sugar GlcNAc to sustain parasite glycoconjugate biosynthesis and carbon metabolism . Studies on the intracellular growth of L . major wild type and mutant lines lacking enzymes in GlcNAc synthesis or catabolism in ex vivo cultured macrophages provided direct evidence that amino sugars generated by breakdown of internalized hyaluronan and other proteoglycans are utilized as carbon sources by amastigotes . First , intracellular survival of L . major ∆gnat promastigotes in macrophages was strongly enhanced by supplementation of the cultures with high molecular weight hyaluronan . In contrast , neither free GlcNAc nor the GlcNAc-rich polymer , chitin , which is mainly degraded extracellularly by macrophages , restored growth . Second , the growth stimulatory activity of high molecular weight hyaluronan was dependent on the expression of the GlcNAc-catabolic pathway in the parasite , as hyaluronan did not rescue the growth defect of the L . major ∆gnd mutant . Third , low molecular weight hyaluronan ( <33 kDa ) , which act as signaling molecule in inflammation , was less effective at rescuing the intracellular growth of L . major ∆gnat parasites than high molecular weight hyaluronan . The effect of hyaluronan on macrophage signaling and Leishmania survival is likely to be negligible , as intracellular survival of wild type L . major was not affected by low molecular weight hyaluronan . Finally , we show that hyaluronan is internalized into the phagolysosome compartment of cultured macrophages and is actively turned over within macrophage-rich skin lesions with a half-life of < 2 days , comparable to previously determined rates of hyaluronan turnover in healthy skin [19 , 23] . Our data suggest that the nutrient environment in cultured macrophages may differ substantively from comparable intracellular niches in lesions . In particular , differences in the extent to which in vitro cultivated and lesion macrophages take up hyaluronan likely accounts for the modest loss of virulence phenotype of the L . major ∆gfat and ∆gnat mutants in mice , but severe loss of intracellular growth in ex vivo macrophages [14] ( and this study ) . Paradoxically , L . major ∆gnd parasites also display reduced intracellular growth in vitro infected macrophages , suggesting that amino sugars remain an important carbon source even in cultured macrophages [13] . Serum that supports growth of cultured macrophages does contain some levels of hyaluronan . It is thus possible that the levels of GlcN/GlcNAc in the phagolysosome of cultured macrophages increase over time , but which is too slow to rescue the highly sensitive ∆gnat and ∆gfat mutants . Together , our results highlight potentially important differences in the nutrient environment of cultured versus lesion macrophages . Leishmania lack detectable hyaluronidase activity and are unable to breakdown and utilize hyaluronan as a carbon source directly . However , Leishmania promastigotes secrete a chitinase and can utilize chitin ( a homopolymer of GlcNAc ) as their sole carbon source in vitro . Overexpression of the L . donovani chitinase resulted in enhanced lesion development [20] , suggesting that this enzyme has a role in virulence . As chitin is not produced by mammals and supplementation of in vitro infected macrophages with chitin failed to restore intracellular growth of ∆gnat , it is possible that the secreted parasite chitinase complements other hydrolases activities in the macrophage phagolysosome involved in the complete break down of hyaluronan . This would potentially enhance the rate of production of free GlcNAc that can then be used by intracellular amastigotes . Hyaluronan degradation in the phagolysosome will also generate free pools of the acidic sugar , glucuronic acid . Leishmania lack enzymes needed to catabolize glucuronic acid and are thus unable to utilize this sugar as sole carbon source ( S7 Fig ) . However , they do contain enzymes needed to convert glucuronic acid to ascorbic acid , an essential vitamin and cofactor for several peroxidases , and this hybrid pathway of de novo synthesis may be important for virulence [26 , 27] . Intriguingly , the saliva of the sandfly vector contains high levels of hyaluronidases , which are injected into the skin together with infective promastigote stages during a natural infection [28 , 29] . The presence of hyaluronidase in the sandfly saliva correlates with enhanced L . major induced skin lesion development in susceptible mice [28] . The sandfly hyaluronidase could potentially enhance promastigotes infectivity in a number of ways; by reducing the barrier properties of the extracellular matrix , stimulating neutrophil recruitment and/or angiogenesis via the generation of bioactive low molecular hyaluronan fragments or by generating hyaluronan fragments that are more readily taken up by Leishmania-infected macrophages . Secretion of sand fly hyaluronidase is also expected to interfere with the wound healing response that occurs after sand fly bites ( and needle puncture ) , as it depends on an orderly regulation of hyaluronan catabolism [19] . There are a number of reports that other pathogens may utilize hyaluronan as a carbon source , including Group A Streptococcus and Mycobacteria tuberculosis [30 , 31] . In contrast to Leishmania , these bacteria secrete hyaluronidases and are therefore able to degrade and use extracellular pools of hyaluronan . In the case of M . tuberculosis , hyaluronan utilization appears to be important during extracellular growth in alveolar spaces , while intracellular stages are dependent on non-carbohydrate carbon sources [31 , 32] . Very few microbial pathogens are able to survive and proliferate long term in the mature phagolysosome of macrophages [9 , 33] . We propose that the dependence of Leishmania parasites on amino sugar metabolism , as well as their requirement for many other nutrients [10] , may have provided the evolutionary drive to colonize this nutritionally diverse niche in the mammalian host .
Use of mice was approved by the Institutional Animal Care and Use Committee of the University of Melbourne ( ethics number 0811011 . 1 ) . All animal experiments were performed in accordance with the Australian National Health Medical Research council ( Australian code of practice for the care and use of animals for scientific purposes , 7th Edition , 2004 , ISBN: 1864962658 ) . Wild type promastigotes of L . major ( MHOM/SU/73/5ASKH ) and L . major Δgnd were cultured in M199 media ( Gibco ) supplemented with 10% heat-inactivated foetal bovine serum ( FBS ) ( Gibco ) or in completely defined media [34] at 27°C . For isolation of transfected parasites , media were supplemented with bleomycin ( 5 μg/ml; Calbiochem ) , nourseothricin ( 70 μg/ml; Sigma ) and G418 ( 100 μg/ml , Calbiochem ) and colonies isolated from M199-agar plates ( 1% , Nu Sieve agarose , FMC BioProducts ) . For the isolation of L . major Δgnat cell line ( S1 Text ) , parasites were additionally supplemented with 50 μg/ml GlcNAc . Metacyclogenesis was determined using peanut agglutinin as described previously [13] . For labeling of L . major wild type and Δgnat parasites with [3H]-Glc ( 1 mCi; Perkin Elmer ) , mid log phase promastigotes ( 108 ) were washed twice in PBS and resuspended in M199 media ( 107 cell/ml ) and incubated for 24 hours at 27°C prior to labeling ( GlcNAc starve ) . Parasites were washed once in PBS and incubated in glucose-free RPMI ( 2 × 108 cell/ml ) for 10 minutes at 27°C ( Glc starve ) before the addition of [3H]-Glc ( 50 μCi/ml ) and incubated for a further 30 minutes at 27°C . Thereafter , parasites were resuspended in PBS and extracted in 300 μl chloroform:methanol:water ( final ratio of 1:2:0 . 8 v/v ) . The extract was partitioned in 1-butanol and water ( 2:1 v/v ) and analyzed by high performance thin layer chromatography ( HPTLC ) [35] . The de-lipidated pellets were analyzed by 12% SDS-PAGE and Western blotting . After transfer onto nitrocellulose membranes ( 0 . 45 μm , Advantec MFS; 100 V , 90 minutes ) , membranes were incubated in blocking buffer ( 5% powdered skim milk in 20 mM Tris-HCl , pH 7 . 6 , 300 mM NaCl , and 0 . 05% Tween 20 ) overnight at 4°C . Membranes were probed with anti-gp63 ( provided by Dr . E . Handman; 1:1000 dilution ) and anti-SMP1 rabbit antibodies ( 1:1000 dilution ) , or the monoclonal anti-phosphoglycan LT15 ( provided by Dr . T . Ilg; 1:1000 dilution ) , all suspended in blocking buffer for 1 hour at RT . Blots were washed in Tris buffer saline containing Tween-20 ( TTBS; 20 mM Tris-HCl , pH 7 . 6 , 300 mM NaCl , 0 . 05% Tween 20 ) for 30 minutes . Secondary antibody ( horseradish peroxidase-conjugated anti-rabbit and anti-mouse secondary antibodies ) were diluted 1:2500 in blocking buffer and applied to blots ( 1 hour , RT ) . After washing in TTBS , binding was detected using ECL reagents ( Amersham ) and analyzed using Gel Pro Analyzer . Log phase L . major wild type and Δgnat promastigotes were washed three times in PBS and resuspended in completely defined medium ( CDM ) ( - hexose , —inositol ) supplemented with 13 mM glucose , +/- 50 μg/ml GlcNAc for 0 , 15 , 30 , 60 and 120 minutes . Promastigotes were hypotonically lysed ( 1 mM NaHEPES pH 7 . 4 , 2 mM EGTA , 2 mM DTT , 40 μl/ml protease inhibitor cocktail [PIC] from Roche Diagnostics ) by chilling on ice for 10 minutes followed by brief sonication . Cell lysates were centrifuged ( 0°C , 2 , 300 x g , 5 minutes ) and resuspended in assay buffer ( 50 mM NaHEPES pH 7 . 4 , 50 mM KCl , 5 mM MgCl2 , 1 mM MnCl2 , 2 mM EGTA , 2 mM DTT , 1 mM ATP ) containing 40 μl/ml PIC . To begin the assay , membranes were incubated with 50 μCi GDP-[3H]-Man ( 0 . 25 mCi; Perkin Elmer ) ( total vol . per assay 80 μl ) for 10 minutes in a 27°C water bath . The reaction was stopped by addition of 300 μl chloroform:methanol ( 1:2 v/v ) and samples then extracted for 2 hours at RT with sonication every 20 minutes . Following biphasic separation in 1-butanol and water ( 2:1 v/v ) , the organic phases were treated with PI-PLC as described in [36] . Samples were then subjected to another round of biphasic separation in 1-butanol and water ( 2:1 v/v ) and the organic phase resuspended in 40% 1-propanol , loaded onto HPTLC plates and developed in chloroform:methanol:NH4OAc:NH3:water ( 180:140:9:9:23 v/v ) before being exposed to film at -70°C . L . major promastigotes were suspended in hypotonic buffer ( 1 mM NaHEPES , pH 7 . 4 , 2 mM EGTA , 2 mM DTT , 40 μl/ml PIC ) and chilled on ice for 10 minutes before being lysed by sonication ( 2 × 4 sec ) . Following lysis , NaHEPES , pH7 . 4 , MgCl2 and acetyl-CoA was added to the lysates to make a final concentration of 50 mM NaHEPES , pH7 . 4 , 5 mM MgCl2 and 150 μM acetyl-CoA . GNAT activity in the lysate ( 3 × 107 cell equivalents in 80 μl ) was measured by addition of GlcN6P ( 1 mM ) and incubating at 27°C for 1 , 10 , 30 or 60 minutes . Controls were either incubated with H2O for 60 minutes at 27°C or boiled for 5 minutes before addition of GlcN6P . In all cases , the reaction was stopped by addition of chloroform:methanol ( 1:2 v/v ) to make a final concentration of 1:2:0 . 8 v/v ( chloroform:methanol:water ) . Samples were extracted ( 2 hours , RT ) , and subjected to biphasic separation in 1-butanol and water ( 2:1 v/v ) as described in 0 . Aqueous fractions containing the hexose phosphates were analyzed by liquid chromatography mass spectrometry on an Agilent 6520 Q-TOF LC/MS Mass Spectrometer coupled to an Agilent 1200 LC system ( Agilent , Palo Alto , CA ) . All data were acquired and reference mass corrected via a dual-spray electrospray ionization ( ESI ) source . Each scan or data point on the Total Ion Chromatogram ( TIC ) is an average of 15 , 000 transients , producing a spectrum every second . Mass spectra were created by averaging the scans across each peak and background subtracted against the first 10 seconds of the TIC . Acquisition was performed using the Agilent Mass Hunter software version B . 02 . 01 and analysis was performed using Mass Hunter version B . 03 . 01 . Mass Spectrometer Conditions: Ionisation mode: Electrospray Ionisation; Drying gas flow: 7 liters/minutes; Nebuliser: 35 psi; Drying gas temperature: 325°C; Capillary Voltage ( Vcap ) : 4000 V; Fragmentor: 100 V; Skimmer: 65 V; OCT RFV: 750 V; Scan range acquired: 100–3000 m/z; Internal Reference ions: Negative Ion Mode = m/z = 112 . 98 and 1033 . 98 . For the isolation of bone-marrow derived macrophages ( BMDMs ) , tibia and femur of BALB/c mice were flushed with RPMI medium 1640 ( Gibco BRL ) supplemented with 15% FBS , 4 mM glutamine ( MultiCel ) , 100 units/ml penicillin , 100 μg/ml streptomycin , and 20% ( v/v ) L929 cell-conditioned media and grown in petri dishes ( 24 hours , 37°C with 5% CO2 ) . BMDMs and RAW 264 . 7 macrophages were grown on 10 mm coverslips in RPMI medium 1640 ( Gibco BRL ) supplemented with 15% FBS , 10% L929 conditioned media , 100 units/ml penicillin and 100 μg/ml streptomycin for 24 hours at 37°C with 5% CO2 . Macrophage monolayers were overlaid with 4-day stationary phase L . major promastigotes ( parasites added to macrophages at a ratio of 10:1 ) and incubated for 4 hours at 33°C with 5% CO2 . Coverslips were washed three times in PBS to remove unattached parasites then incubated in fresh macrophage medium ( RPMI + 15% FBS ) for up to 6 days at 33°C in 5% CO2 . Chitin ( crab shells , Sigma ) , high molecular weight hyaluronan ( Streptococcus isolate , Sigma ) or polymers with defined molecular weight ( obtained after acid hydrolysis and high performance liquid chromatography and size exclusion chromatography and multiple-angle laser light scattering [HPLC SEC-MALLS] ) were added to macrophage cultures at 10 μg/ml . Coverslips were washed in PBS to remove unattached parasites and sequentially incubated in methanol ( 25°C , 10 minutes ) , PBS containing 50 mM NH4Cl ( 25°C , 10 minutes ) , and 1% BSA in PBS ( 25°C , 30 minutes ) . The fixed cells were probed with mAb anti-LAMP ( 1:100 dilution in 1% BSA in PBS; BD Biosciences ) and Alexa Fluor-488 goat anti-rat ( 1:1000 dilution; Molecular Probes ) to visualize PV membranes . Infected macrophages were incubated with 0 . 2 μg/ml propidium iodide ( Sigma ) and mounted in Mowiol 4–88 ( 5 μl; Calbiochem ) containing Hoechst 33342 ( 8 μg/ml; Life Technologies ) to visualize the plasma membrane , nuclei and kinetoplasts , respectively . Hyaluronan was detected by live-cell microscopy by covalently linking high molecular weight polymers to fluorescein ( FITC ) [37] . Macrophages were infected with L . major or L . mexicana ( M379 ) for 24 hours before adding hyaluronan-FITC for an additional 18 hours and LysoTracker-Red DND-99 ( Invitrogen ) for 60 minutes . Fluorescence microscopy was performed using a Zeiss Axioplan2 imaging microscope , equipped with AxiCam MRm camera and the AXIOVISION 4 . 3 software ( Zeiss ) . Images were compiled in Photoshop Elements . Female BALB/c mice ( 6–8 weeks old ) were maintained in a pathogen-free facility ( Bio21 Institute , University of Melbourne ) . Groups of mice ( five per treatment ) were inoculated intradermally at the base of the tail with stationary-phase promastigotes ( 106 in 50μl sterile PBS ) or lesion-derived amastigotes ( 2 × 105 parasites in 50 μl sterile PBS ) . The size of the developing lesions was monitored and recorded weekly as described in [38] . Data is expressed as arithmetic mean ( ± standard error mean ) of the lesion scores for the groups of five mice . For amastigote infections , excised lesions ( 5–10 mm ) from BALB/c mice were placed in chilled 1 × PBS and pushed through a plastic sieve using the flat end of a 5 ml syringe plunger and syringed three times using a 27 ½ gauge needle . Amastigotes were separated from cell debris by centrifugation ( 30 x g , 10 minutes , 4°C ) and the pellet washed twice in chilled PBS ( 850 x g , 10 minutes , 4°C ) . Amastigote numbers were determined using a haemocytometer . L . mexicana infected BALB/c mice containing skin lesions of score of 2 were given an intraperitoneal bolus of 35 μl/g bodyweight 2H2O ( Cambridge Isotope Laboratories ) and 0 . 9% NaCl and subsequently fed with 9% 2H2O in their drinking water for up to 5 weeks . This resulted in the stable enrichment of 5% 2H2O of total body water as determined by [39] . Excised , lesions were cut into small fragments ( about 1mm3 ) and digested with hyaluronidase ( Type I; Sigma ) for 18 hours at 37°C . After biphasic extraction with chloroform and methanol ( 1:2 v/v ) , the released tetrasaccharides in the aqueous phase were analyzed by liquid chromatography and mass spectrometry ( LC-MS ) using an LC-QTOF-MS—an Agilent 1290 LC-system coupled to an Agilent 6550 Electrospray Ionisation-Quadrupole Time of Flight ( Agilent Technologies , Singapore ) . Tandem MS was used to distinguish hyaluronic acid ( HA ) from N-acetylheparosan [40] . Amino sugar species were separated by injecting 10 μL of samples into Phenomenex Hyperclone C18 column –3 . 8 μm 4 . 6 x 100 mm with a flow rate of 200 μL/minutes . The mobile phases used were 10 mM ammonium acetate ( Sigma Aldrich ) in water ( A ) and then 10 mM ammonium acetate in acetonitrile ( B ) . The column temperature was kept at 20°C with a 10 minute run time . The chromatographic separation of the amino sugars was achieved by a gradient of mobile phase B increased from 1% to 50% during the first 6 minutes , held for two minutes followed by equilibrating the column for another two minutes as the initial composition . Mass spectra for the ions were collected in negative ionization mode . The full scan spectra ( 100–1700 Da ) was collected using the following conditions as fragmentor: voltage 175 V , nebulizer pressure 45 psi and capillary voltage 4000 V . Nitrogen was used as a drying ( 13 L/minutes ) and sheath gas flow as 12 L/minutes with sheath gas temperature 275°C . Total ion chromatograms ( TIC ) and mass spectra were processed using an Agilent MassHunter software , version 6 . The ion 775 . 2 was chosen as a diagnostic ion following the analysis of digested HA standards . The M0 , M1 , M2 , M3 , and M4 were quantitated . The half-life was determined by plotting the fraction of new HA ( excess molar enrichment divided by the maximal excess molar enrichment ( 4 weeks labeled lesion ) ) over time . | Macrophages are the primary host cells for a number of important microbial pathogens , including protozoan parasites belonging to the genus Leishmania . With few exceptions , little is known about the nutrient composition of the vacuolar compartments occupied by these pathogens . Leishmania proliferate within the mature phagolysosome compartment of macrophages and recent studies have suggested that intracellular parasite stages are dependent on the uptake of amino sugars . However , how Leishmania gain access to these sugars is unclear . In this study we have generated a Leishmania major mutant that is a strict auxotroph for the amino sugar , N-acetylglucosamine ( GlcNAc ) . This mutant exhibited a similar virulence phenotype as wild type parasites in infected mice , but was unable to survive in cultured macrophages . The intracellular survival of the GlcNAc-auxotroph in cultured macrophages was restored by supplementation of the medium with the high molecular weight glycosaminoglycan , hyaluronan , which is rich in GlcNAc . Hyaluronan is a major component of vertebrate extracellular matrix and we show that it is rapidly degraded in Leishmania-induced skin lesions . Hyaluronan is internalized by infected macrophages and traffics to the Leishmania containing phagolysosome . Leishmania thus appear to exploit the critical role of macrophages in extracellular matrix turnover to obtain essential sugar carbon sources for growth and virulence . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Intracellular Survival of Leishmania major Depends on Uptake and Degradation of Extracellular Matrix Glycosaminoglycans by Macrophages |
The spider family Sicariidae includes two genera , Sicarius and Loxosceles . Bites by Sicarius are uncommon in humans and , in Brazil , a single report is known of a 17-year old man bitten by a Sicarius species that developed a necrotic lesion similar to that caused by Loxosceles . Envenomation by Loxosceles spiders can result in dermonecrosis and severe ulceration . Sicarius and Loxosceles spider venoms share a common characteristic , i . e . , the presence of Sphingomyelinases D ( SMase D ) . We have previously shown that Loxosceles SMase D is the enzyme responsible for the main pathological effects of the venom . Recently , it was demonstrated that Sicarius species from Africa , like Loxosceles spiders from the Americas , present high venom SMase D activity . However , despite the presence of SMase D like proteins in venoms of several New World Sicarius species , they had reduced or no detectable SMase D activity . In order to contribute to a better understanding about the toxicity of New World Sicarius venoms , the aim of this study was to characterize the toxic properties of male and female venoms from the Brazilian Sicarius ornatus spider and compare these with venoms from Loxosceles species of medical importance in Brazil . SDS-PAGE analysis showed variations in the composition of Loxosceles spp . and Sicarius ornatus venoms . Differences in the electrophoretic profiles of male and female venoms were also observed , indicating a possible intraspecific variation in the composition of the venom of Sicarius spider . The major component in all tested venoms had a Mr of 32–35 kDa , which was recognized by antiserum raised against Loxosceles SMases D . Moreover , male and female Sicarius ornatus spiders' venoms were able to hydrolyze sphingomyelin , thus showing an enzymatic activity similar to that determined for Loxosceles venoms . Sicarius ornatus venoms , as well as Loxosceles venoms , were able to render erythrocytes susceptible to lysis by autologous serum and to induce a significant loss of human keratinocyte cell viability; the female Sicarius ornatus venom was more efficient than male . We show here , for the first time , that the Brazilian Sicarius ornatus spider contains active Sphingomyelinase D and is able to cause haemolysis and keratinocyte cell death similar to the South American Loxosceles species , harmful effects that are associated with the presence of active SMases D . These results may suggest that envenomation by this Sicarius spider has the potential to cause similar pathological events as that caused by Loxosceles envenomation . Our results also suggest that , in addition to the interspecific differences , intraspecific variations in the venoms composition may play a role in the toxic potential of the New World Sicarius venoms species .
The spider family Sicariidae includes two genera , Sicarius and Loxosceles . Sicarius species ( six-eyed crab spiders , six-eyed sand spiders ) live in dry forests and deserts throughout Southern Africa , South America and Central America . The genus Sicarius is composed of robust flattened spiders , 9–19 mm long and a leg span of about 5 cm . The legs are laterally placed , resembling a crab , hence the common name . These spiders are found buried in soil layers , where they live and wait to trap their prey ( Figure 1 ) . They feed on passing insects , rapidly emerging from the sand when disturbed . During self-burial , soil particles can adhere to their specialized sethae ( hairs ) , which cover their bodies , changing their natural coloration to the color of the environment [1] , [2] . Bites by Sicarius are uncommon in humans . In Brazil , a single report is known of a 17-year old man bitten by a Sicarius species who developed necrotic lesion similar to that caused by Loxosceles [3] . Yet , several reports have shown that African Sicarius spider venoms are lethal to rabbits and are also able to cause dermonecrotic lesions in humans and in experimental animals [4]–[9] . The genus Loxosceles includes spiders of small dimensions , with a body length of about 10 mm and leg span of 3 cm , relatively fine long legs and six eyes organized in a characteristic pattern of three dyads in the shape of a U . Known popularly as brown spider in South America and as recluse spider in North America , the specimens have a dark or light brown violin-shaped mark on its carapace . Loxosceles spiders are nocturnal and prefer dry , dark and quiet places , and live under wood and rocks , under the bark of trees and in caves . They adapt very well to domiciliary conditions , hiding behind pictures , in furniture , in clothes and shoes , always protected from direct light [10]–[12] . Envenomation by Loxosceles spiders , considered one of the four dangerous forms of araneism [13] , is a serious public health hazard in North and South America . Although systemic reactions such as shock , haemolysis , renal insufficiency and disseminated intravascular coagulation are rare , small areas of erythema often leading to larger areas of ulceration and necrosis are frequently observed [14] , [15] . At least three different synanthropic Loxosceles species of medical importance are known in Brazil ( L . intermedia , L . gaucho , L . laeta ) and more than 5000 cases of envenomation by these spiders are reported each year . Sicarius and Loxosceles spider venoms share a common characteristic , i . e . , the presence of Sphingomyelinases D ( SMase D ) [16]–[17] . Numerous studies have demonstrated that SMase D present in the venoms of Loxosceles spiders is the main component responsible for the local and systemic effects observed in loxoscelism [18]–[26] . SMases D hydrolyze sphingomyelin resulting in the formation of ceramide-1-phosphate and choline [18] , [19] , [21] and , in the presence of Mg2+ , are able to catalyze the release of choline from lysophosphatidylcholine [27] . Recently , it was demonstrated that Sicarius species from Africa , like Loxosceles spiders from the Americas , present high venom SMase D activity . However , despite the presence of SMase D like proteins in venoms of several New World Sicarius species tested , as the ones from Argentina ( S . terrosus , S . rupestris , S . patagonicus ) , Peru ( S . peruensis ) and Costa Rica ( S . rugosus ) , these venoms had reduced or not detectable SMase D activity [17] . In order to contribute to a better understanding about the toxicity of New World Sicarius venoms , the aim of this study was to characterize the biochemical and biological properties of male and female venoms from a Brazilian Sicarius species , Sicarius ornatus , and compare these with venoms from Loxosceles species of medical importance in Brazil .
Tween 20 , bovine serum albumin ( BSA ) , paraformaldehyde , 3- ( 4 , 5 dimethylthiazol-2yl ) -2 , 5 diphenyltetrazolium bromide ( MTT ) , sphingomyelin ( SM ) , choline oxidase , horseradish peroxidase ( HRPO ) and 3- ( 4-hydroxy-phenyl ) propionic acid were purchased from Sigma Co . ( St . Louis , MO , USA ) . 5-bromo-4-chloro-3-indolyl-phosphate ( BCIP ) , nitroblue tetrazolium ( NBT ) and goat anti-rabbit IgG-alkaline phosphatase ( GAR/IgG-AP ) were from Promega Corp . ( Madison , WI , USA ) . Rabbit anti-mouse IgG-FITC ( RAM-FITC ) and goat anti-rabbit IgG-FITC ( GAR-FITC ) were from Amersham Pharmacia Biotech ( Buckinghamshire , UK ) . Monoclonal antibody against GPC ( Bric4 , extracellular epitope aa 16–23 ) was from IBGRL ( Bristol , UK ) . Rabbit serum against SMases D from L . intermedia venom was obtained as previously described [20] . Buffers were: veronal-buffered saline ( VBS2+ ) , pH 7 . 4: 10 mM NaBarbitone , 0 . 15 mM CaCl2 and 0 . 5 mM MgCl2; phosphate-buffered saline ( PBS ) , pH 7 . 2: 10 mM NaPhosphate , 150 mM NaCl; fluorescence activated cell sorter ( FACS ) buffer , containing PBS , 1% BSA , 0·01% sodium azide . HEPES-buffered saline ( HBS ) , pH 7 . 4: 10 mM Hepes: 140 mM NaCl , 5 mM KCl , 1 mM CaCl2 , 1 mM MgCl2 . Adult male ( n = 3 ) and female ( n = 5 ) Sicarius ornatus spiders ( Figure 1 ) were collected in Elisio Medrado , State of Bahia , RPPN Jequitiba ( capture and maintenance licenses from IBAMA , Brazil , number 13676-1 ) . Eleven voucher specimens were deposited at Arachnology Laboratory , Museu Nacional do Rio de Janeiro ( MNRJ ) under accession numbers 06479 to 06485 . Adult females of Loxosceles laeta , L . gaucho and L . intermedia were provided by Immunochemistry Laboratory , Butantan Institute , Brazil ( capture and maintenance licenses from IBMA , Brazil , number 11971-2 ) . We considered as adults spiders those specimens with fully developed palpal copulatory organs ( males ) or with an epigastric furrow with a clearly visible opening of the oviduct ( females ) . The venoms were obtained by electrostimulation by the method of Bucherl [28] with slight modifications . Briefly , 15–20 V electrical stimuli were repeatedly applied to the spider sternum and the venom drops were collected with a micropipette in PBS , aliquoted and stored at −20°C . The protein content of the samples was evaluated using the BCA Protein Assay Kit ( Pierce Biotechnology , MA , USA ) . Venom samples ( 10 µg of protein ) from Sicarius ornatus ( male and female ) or Loxosceles spp . ( female ) were solubilised in non-reducing sample buffer , run on 12% SDS-PAGE [29] and silver stained . Alternatively , gels were blotted onto nitrocellulose [30] . After transfer , the membranes were blocked with PBS containing 5% BSA and incubated with rabbit serum anti-native SMases D from L . intermedia venom ( diluted 1∶250 ) for 1 h at room temperature . Membranes were washed three times with PBS/0 . 05% Tween 20 for 5 min each wash , and incubated with GAR/IgG-AP ( 1/7500 ) in PBS/1% BSA for 1 h at room temperature . After washing three times with PBS/0 . 05% Tween 20 , for 5 min each wash , blots were developed using NBT/BCIP according to the manufacturer's instructions ( Promega ) . The SMase D activity of the venoms was estimated by determining the choline liberated from lipid substrates , using a fluorimetric assay [31] . Briefly , sphingomyelin ( SM – 50 µM ) was diluted in 1 mL HEPES-buffered saline ( HBS ) , samples of Sicarius ornatus or Loxosceles spp . venoms ( 10 µg of protein ) were added and the reaction was developed for 30 min at 37°C . After incubation , a mixture consisting of 1 unit choline oxidase/mL , 0 . 06 units of horseradish peroxidase/mL and 50 µM of 3- ( 4-hydroxy-phenyl ) propionic acid in HBS was added and incubated for 10 min . The choline liberated was oxidized to betaine and H2O2 and this product determined by fluorimetry at λem = 405 nm and λex = 320 nm , using 96-well microtiter plates , in a spectrofluorimeter ( Perkin-Elmer , USA ) . Human blood was obtained from healthy donors who knew the objectives of the study and signed the corresponding informed consent form approved by the ethics committee ( CAAE: 07039213 . 3 . 0000 . 5467 ) . Blood samples were collected without anticoagulant and allowed to clot for 4 hours at 4°C . After centrifugation , normal human serum ( NHS ) was collected and stored at −80°C . Blood samples drawn to obtain erythrocytes ( E ) for subsequent use as target cells were collected in anticoagulant ( Alsever's old solution: 114 mM citrate , 27 mM glucose , 72 mM NaCl , pH 6 . 1 ) . Human erythrocytes were washed and resuspended at 2% in VBS2+ and incubated with different concentrations of the venoms for 1 h at 37°C . Control samples were incubated with VBS2+ . The cells were washed , resuspended to the original volume in VBS2+ and analysed in a haemolysis assay as described [20] or prepared for flow cytometry . Samples of human erythrocytes ( 25 µL ) were incubated for 30 min with 25 µL of primary or control antibodies ( 1–10 µg/mL ) in FACS buffer . After washing , cells were incubated with the appropriate FITC-labelled secondary antibodies for 30 min . The cells were washed and fixed in FACS buffer containing 1% paraformaldehyde and analysed by flow cytometry ( FACScalibur , Becton Dickinson , California , USA ) . Human keratinocytes ( cell line HaCaT ) were maintained in DMEM ( Gibco-BRL , Gaithersburg , MD , USA ) , supplemented with 10% ( vol/vol ) heat-inactivated ( 56°C , 30 min ) foetal bovine serum ( FBS; Cultilab , São Paulo , Brazil ) , 100 IU of penicillin/mL , and 100 IU of streptomycin/mL at 37°C in humidified air with 5% CO2 . HaCaT cells were subcultured in 96-well plates ( 5×104cells/well ) . Cells at 50%–70% confluence were maintained overnight in DMEM without FBS , followed by incubation with the venoms ( 10 µg of protein ) . DMEM without FBS was used as the control . After 48 and 72 hours , the viability of the cultures was tested by the MTT [32] . Supernatants of each sample ( 100 µL ) were collected and mixed with 100 µL of water and the absorbance was measured in a spectrophotometer ( Multiskan-EX , Labsystems , Helsinki , Finland ) at 540 and 620 nm . The relative cell viability was calculated as: [ ( Sample OD ( 540–620 nm ) – Background control OD ( 540–620 nm ) ) / ( Control OD ( 540–620 nm ) – Background OD ( 540–620 nm ) ) ]×100 . Data were analyzed statistically by one way ANOVA and Tukey test . A P-value<0 . 05 was considered significant .
Figure 2 shows that the venom of male and female Sicarius ornatus spiders contain similar amounts of protein . Comparison analysis showed that male and female Sicarius ornatus venoms contain significant higher protein concentrations than L . laeta female venom . Comparative analysis of the spider venoms , by SDS-PAGE followed by silver staining , revealed differences in the number and intensity of bands of venoms from male and female Sicarius ornatus spiders and also from Loxosceles species , however , all venoms showed a major band with Mr of 32–35 kDa , which corresponds , in Loxosceles venoms , to the main toxic component , i . e , the SMase D ( Figure 3A ) . In order to assess the identity of this band and analyze the inter- and intra-species cross-reactivities , polyclonal antiserum raised against a pool of purified SMases D from L . intermedia was used in western blot . Figure 3B shows that this antiserum strongly recognized the SMases D present in the venoms from Loxosceles intermedia , L . laeta and L . gaucho and also reacted with a band of similar Mr of approximately , 33 kDa in the Sicarius ornatus spider male and female venoms , suggesting that this band also corresponds to a sphingomyelinase D . Figure 4 shows that all tested venoms , including from the Sicarius ornatus spider , were able to hydrolyze sphingomyelin . Comparative analysis revealed significant differences in the sphingomyelinase activity of venoms from Loxosceles species and Sicarius ornatus However , Loxosceles spp . and Sicarius ornatus female venoms exhibited a more potent sphingomyelinase activity than Sicarius ornatus male venoms ( P<0 . 005 ) . To compare the spiders' venoms capability of inducing complement-dependent haemolysis , human erythrocytes were incubated Sicarius ornatus venoms or Loxosceles spp . venoms and incubated with normal human serum as a source of complement . Figure 5 shows that male and female venoms from Sicarius ornatus spiders , as well as from Loxosceles , were able to render human erythrocytes susceptible to lysis by autologous serum . A more potent complement-dependent haemolytic inducing activity was detected in Loxosceles spp . venoms ( P<0 . 005 ) . We have previously shown that removal of glycophorins ( GPs ) is partially responsible for the increased complement-susceptibility of erythrocytes treated with Loxosceles venoms . To investigate if Sicarius ornatus venoms induced a similar effect , human erythrocytes were incubated with Sicarius ornatus or Loxosceles venoms or buffer and analyzed for the expression of glycophorin C ( GPC ) by flow cytometry . A significant reduction in binding of anti-GPC antibodies was observed after treatment of erythrocytes with all venoms ( #Figure 6A ) . Statistically significant differences ( P<0 . 005 ) were detected between Loxosceles and Sicarius ornatus venoms and also between Sicarius ornatus genders , the female venom being more potent inducer of removal of GPs . The disappearance of GPs epitopes , induced by both Sicarius ornatus and Loxosceles venoms , was associated with the binding of the SMAses D to the erythrocyte cell surface as detected by anti-Loxosceles SMAse D anti-serum ( Figure 6B ) . Nonetheless , the stronger reduction in the expression of GPs in Loxosceles venom treated cells , as compared with male and female Sicarius ornatus venoms , correlates positively with the higher SMase D cell binding and hemolytic complement-dependent inducing capabilities exhibited by the former venom . Envenomation by Loxosceles spiders is a well-documented cause of necrotic skin lesions in humans . Using HaCaT cultures , a human keratinocyte cell line , as an in vitro model for cutaneous loxoscelism , we have shown that Loxosceles spider venom and its SMase D induce apoptosis in human keratinocytes . In order to analyze if the same toxic effect could be induced by Sicarius ornatus venoms , HaCaT cells were incubated with male or female Sicarius ornatus venoms or L . laeta venom during 48 or 72 h and the cell viability was analyzed by the MTT method . Figure 7 shows that both female Loxosceles and Sicarius ornatus venoms were able to induce a significant loss of cell viability after 72 h of incubation; the female Sicarius ornatus venom was more efficient in provoking the loss of cell viability than male Sicarius ornatus and female Loxosceles venoms .
In the present study , we have investigated the toxic potential of venoms from male and female Brazilian Sicarius ornatus spider , collected in Elisio Medrado , Bahia , Brazil and compared them with the venoms from Loxosceles species of medical importance in Brazil . We show here that this Brazilian Sicarius ornatus venom is endowed with all toxic biological properties ascribed to the venoms from Loxosceles species , including the abilities to hydrolyze sphingomyelin , to induce keratinocyte cell death and complement dependent haemolysis . SDS-PAGE analysis showed variations in the composition of Loxosceles spp . and Sicarius ornatus venoms . Differences in the electrophoretic profiles of male and female venoms were also observed , indicating a possible intraspecific variation in the composition of the venom of this Brazilian Sicarius sp . spider , as we have previously described for Loxosceles venoms [33] , [34] . Interestingly , the major component in all tested venoms had a Mr of 32–35 kDa , which corresponds in Loxosceles venoms to the main toxic components , the SMases D . This major band , in the Sicarius ornatus spider venoms , was also recognized by a monospecific polyclonal serum elicited against Loxosceles SMases D , confirming the presence of SMase D related proteins in male and female venoms from Sicarius ornatus . It has been shown that the SMase D activity from African Sicarius venoms was similar to that of Loxosceles from the Americas; however , little or no activity , at the same venom concentrations , was detected in samples from several South and Central American Sicarius species [16] . Nevertheless , our results showed here indicate that venoms of a Brazilian Sicarius , S . ornatus , is able to hydrolyze sphingomyelin and that the enzymatic activity was similar to that determined for Loxosceles venoms . However , the SMase D activity of Sicarius ornatus male venom was statistically lower than female and Loxosceles spp . venoms , which reinforce the idea of intraspecific variation in the composition and toxicity of the Brazilian Sicarius ornatus venoms . The discrepancy between our data and results from the study of Binford and collaborators [16] are most likely due to interspecies variations of South American Sicarius venoms . Besides , it was not mentioned if the New World Sicarius venoms tested were collected from male or female spiders [16] , thus it is possible to consider that the low SMase D activity detected may be also due to the presence of high amounts of male venoms , with lower activity , in the samples used in the experiments . The venom of Sicarius albospinosus from South Africa can induce systemic effects , including disseminated intravascular coagulation in rabbits , but the same effect was not observed with Sicarius testaceus ( South Africa ) venom [5] , [9] . Thus , these data also suggest that interspecific variations in the venom composition may contribute to the severity of the Sicarius envenomation . Although some studies have investigated the expression and the SMase D activity in Sicarius spider venoms , none of them has addressed the important clinical manifestation induced by Loxosceles SMases D , namely , hemolysis . Investigations focusing on the effects of Loxosceles venoms on erythrocytes demonstrated that SMase D induced activation of membrane bound metalloproteinases , resulting in cleavage of glycophorins , which facilitated activation of complement via the alternative pathway resulting in lysis of the cells [22] . In order to assess whether the Sicarius ornatus could also induce Complement-dependent hemolysis , erythrocytes were incubated with male or female crude venoms . Both venoms were able to render human erythrocytes susceptible to lysis by autologous complement , although with less potency than Loxosceles spp . venoms . Moreover , as with Loxosceles spp . venoms , male and female Sicarius ornatus venoms were able to significantly reduce the binding of anti-GPC antibodies , indicating the recognition of extracellular epitopes close to the membrane . Again , the female venom was more active than male venom . The ability of the Loxosceles SMase D to bind to different species of erythrocytes ( such as human , sheep , rats , rabbits and guinea pigs ) as well as to several cell types ( such as epidermal cells , hepatocytes , monocytes , B and T cells , endothelial cells , platelets and neutrophils ) have been already described [15] , [35] . Although no specific receptor has been described yet for this interaction , the SMase D binding ability certainly is an important step for the mechanism of action of the Loxosceles spider venoms . As shown here , the reduction of GPs epitopes , induced by both Sicarius ornatus and Loxosceles venoms , correlated with the binding of the SMases D to the erythrocyte cell surface as detected by anti-SMase D serum . Although , the data obtained also suggest that the SMases D from Loxosceles can bind to the cells membrane with higher efficiency than Sicarius ornatus ones , but this may be in part a result of differences in the antigenic recognition , since the anti-SMase D serum used was produced against a pool of native Loxosceles SMases D . Together , these observations indicate that Sicarius ornatus venoms also have the ability to induce complement-dependent hemolysis , which may occur by the same hemolytic molecular mechanism displayed by Loxosceles venoms . Several reports have shown that Sicarius as well as Loxosceles venoms are able to cause dermonecrotic lesions in humans and in experimental animals [3] , [9] , [21] , [23] , [25] , [26] , [34] . We have previously demonstrated that Loxosceles spider venom induces an increase in cell death in the keratinocytic human cell line HaCaT [26] . Here , when the HaCaT cells were incubated with Loxosceles or Sicarius ornatus female venoms , a significant decrease in the cell viability was observed , suggesting that Sicarius ornatus venoms may also share similar molecular mechanisms leading to tissue damage and development of dermonecrosis with the Loxosceles venoms . After 72 h , Sicarius ornatus male venom has only induced a small reduction in the cell viability , data that once more strengthens the idea of intraspecific variation in Sicarius ornatus venom toxicity . The lower cytotoxic effect exhibited by male venom did not correlate well to its significant sphingomyelinase activity . The method used for measuring and comparing the sphingomyelinase activity of the venoms is based on the hydrolysis of the substrate sphingomyelin dispersed in a buffer , which maybe does not reflect the in vivo lipase activity on real substrates , e . g . , sphingomyelin present on intact cell membranes . Finally , if the amount of protein and volume of the venom in the poisonous gland is taken into account , the toxic potential of Sicarius ornatus bite is greater than that of Loxosceles , since Sicarius ornatus contains higher volume ( data not shown ) and higher amount of protein in its venom gland than L . laeta spider , whose venom contains the highest protein concentration in Loxosceles species of medical importance in Brazil [33] , [34] . Yet , the reason why few incidences of envenomation by Sicarius in Brazil are reported may lie on the differences in habitat and on the low exposure to humans by the Sicarius species . In conclusion , we show here , for the first time , that a Brazilian Sicarius spider species , Sicarius ornatus , is able to cause haemolysis and keratinocyte cell death similar to the South American Loxosceles species , harmful effects that were positively associated with the presence of active SMases D and with in vivo pathologies . Therefore the venom of S . ornatus has the potential to cause serious pathology upon envenomation , similar to that observed after Loxosceles envenomation . Our results also suggest that , in addition to the interspecific differences , intraspecific variations in the venoms composition may play a role in the toxic potential of the New World Sicarius venoms species . | The spider family Sicariidae includes two genera , Sicarius and Loxosceles . These spiders' venoms share a common characteristic , i . e . , the presence of Sphingomyelinases D ( SMase D ) . This toxin is the main component responsible for the local and systemic effects observed in loxoscelism . In the present study , we have investigated the toxic potential of male and female Brazilian Sicarius ornatus spider venoms and compared these with the venoms from Loxosceles species of medical importance in Brazil . We show here that Brazilian Sicarius ornatus venom is endowed with all toxic in vitro and ex vivo biological properties ascribed to the venoms from Loxosceles species , including the abilities to hydrolyze sphingomyelin and to induce keratinocyte cell death and complement dependent haemolysis , detrimental effects that were positively associated with the presence of active SMases D and with in vivo pathologies . Therefore , the venom of Sicarius ornatus spider can potentially lead to a similar pathology as that observed for Loxosceles envenomation . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"zoology",
"biochemistry",
"immunology",
"biology"
] | 2013 | Venom of the Brazilian Spider Sicarius ornatus (Araneae, Sicariidae) Contains Active Sphingomyelinase D: Potential for Toxicity after Envenomation |
Flea-borne diseases have a wide distribution in the world . Studies on the identity , abundance , distribution and seasonality of the potential vectors of pathogenic agents ( e . g . Yersinia pestis , Francisella tularensis , and Rickettsia felis ) are necessary tools for controlling and preventing such diseases outbreaks . The improvements of diagnostic tools are partly responsible for an easier detection of otherwise unnoticed agents in the ectoparasitic fauna and as such a good taxonomical knowledge of the potential vectors is crucial . The aims of this study were to make an exhaustive inventory of the literature on the fleas ( Siphonaptera ) and range of associated hosts in Iran , present their known distribution , and discuss their medical importance . The data were obtained by an extensive literature review related to medically significant fleas in Iran published before 31st August 2016 . The flea-host specificity was then determined using a family and subfamily-oriented criteria to further realize and quantify the shared and exclusive vertebrate hosts of fleas among Iran fleas . The locations sampled and reported in the literature were primarily from human habitation , livestock farms , poultry , and rodents’ burrows of the 31 provinces of the country . The flea fauna were dominated by seven families , namely the Ceratophyllidae , Leptopsyllidae , Pulicidae , Ctenophthalmidae , Coptopsyllidae , Ischnopsyllidae and Vermipsyllidae . The hosts associated with Iran fleas ranged from the small and large mammals to the birds . Pulicidae were associated with 73% ( 56/77 ) of identified host species . Flea-host association analysis indicates that rodents are the common hosts of 5 flea families but some sampling bias results in the reduced number of bird host sampled . Analyses of flea-host relationships at the subfamily level showed that most vertebrates hosted fleas belgonging to 3 subfamilies namely Xenopsyllinae ( n = 43 ) , Ctenophthalminae ( n = 20 ) and Amphipsyllinae ( n = 17 ) . Meriones persicus was infested by 11 flea subfamilies in the arid , rocky , mountainous regions and Xenopsyllinae were hosted by at least 43 mammal species . These findings place the Persian jird ( M . persicus ) and the Xenopsyllinae as the major vertebrate and vector hosts of flea-borne diseases in Iran including Yersinia pestis , the etiological agent of plague . We found records of at least seven vector-borne pathogenic agents that can potentially be transmitted by the 117 flea species ( or subspecies ) of Iran . Herein , we performed a thorough inventary of the flea species and their associated hosts , their medical importance and geographic distribution throughout Iran . This exercise allowed assessing the diversity of flea species with the potential flea-borne agents transmission risk in the country by arranging published data on flea-host associations . This information is a first step for issuing public health policies and rodent-flea control campaigns in Iran as well as those interested in the ecology/epidemiology of flea-borne disease .
Vector-borne diseases ( VBDs ) are globally responsible for more than 17% of all infectious diseases [1] . There are a large number of viral , rickettsial , bacterial and parasitic diseases that are transmitted by insect vectors [2] . In the last two decades , many zoonotic VBDs have emerged in areas where they previously did not occur , and the incidence of these diseases both in endemic areas and outside their known range has increased [3] . In recent years , most studies on zoonotic diseases have focused on tick- and mosquito-borne diseases , less attention has been given to flea-borne diseases[4] . Fleas ( Siphonaptera ) are small , bloodsucking or hematophagous ectoparasites that may transmit pathogens through several possible mechanisms , including: contaminated feces ( e . g . R . typhi , B . henselae ) , soiled mouthparts ( e . g . Y . pestis , viral pathogens ) , regurgitation of gut contents ( e . g . Y . pestis ) , and infectious saliva ( e . g . R . felis in salivary glands ) [4] . Over 2500 flea species belonging to 16 families and 238 genera have been described worldwide [5] . Fleas are mainly ectoparasites of mammals while birds are infested by only 6% of the known species . This is partly due to reduced collection efforts and sampling bias as only few bird fleas are in close contact with humans [6] . Fleas are one of the most common insect groups that can serve as vector and intermediate host of pathogenic zoonotic agents between vertebrate hosts , including humans [4 , 7–8] . Fleas can have a direct pathogenic effect by causing allergic dermatitis [9–10] or paralysis subsequent to the injection of saliva into their hosts skin or blood [11] . Notorious human pathogens such as Yersinia pestis ( plague ) , Rickettsia typhi ( murine typhus ) , Francisella tularensis ( tularemia ) and Bartonella henselae ( cat scratch disease ) are transmitted by fleas [12–15] . Some fleas tend to be host specific ( restricted or specialist ) , but others have a wide host range ( permissive , opportunistic ) . The permissive species group are more significant than the restricted ones , because they can spread infectious agents among and within their multiple hosts and across a diverse series of habitats [6] . In order to prevent or control the occurrence and spread of flea-borne diseases , it is thus necessary to establish a taxonomical inventory of the flea fauna and their specific distribution range . Climate changes , due to global warming and human intervention , have led to changes in the biological parameters and distribution ranges of vectors and hence of VBDs [16] . On the bases of vulnerability assessments and models , it is predicted that climate change will result in raised incidence of communicable diseases embracing VBDs; however the short and long term effects will be mitigated and will be linked to vector life cycles ( e . g . : developments of preimaginal stages ) and geographic area [17] . Reasonable proofs tend to suggest that changes in climatic factors may affect VBDs incidence especially acting on the off-host developmental life stages of arthropods and hence disease transmission dynamics . Insects as poikilotherm organisms have no internal control of their body temperatures , and as such depend on their host ( s ) —the imago as a transient habitat - , and abiotic conditions for survival , which both condition their vector capacity , as well as their reproduction rate[18] . Moreover , vector capacity is linked to the nature of the pathogen transmitted , survival rate inside its vector host—which may or may not affect vector fitness—and incubation or turnover rate that is inversely proportional to temperature[19] . Moreover , climate and human behavior changes increase human exposures to vectors and the pathogenic agents they transmit [20] . Studies of plague transmission in the U . S . A , China and Kazakhstan have found that the patterns of human or rodent plague are shifting as temperatures warms up or link to climatic oscillations ( such as El Niño ) and precipitation pattern [20] . Iranian physicians were familiar with the human plague for a long time . Although there are little information about the situation of plague from earlier centuries , more documented evidence are available from the 19th and 20th centuries . As a matter of fact , faunistic studies of Iranian fleas have been carried out mainly about 60 years ago in a context of plague research and most species described at the time were collected and described off plague hosts [21] . When plague research stopped , flea inventories did so too and there are no current updates on the flea fauna of Iran . However , a recent study detected antibodies against Y . pestis in dogs—known to be a good sentinels for plague surveillance- while human plague hasn’t been reported for 50 years [22] . This finding triggered some concern about the possible plague reemergence in the countryside , in the old plague foci and called for an update on the state-of-knowledge of the flea diversity in the country . The aims of the present study were to update by reviewing the current state of knowledge of the Iranian Siphonaptera diversity , their host range and especially the medically important species .
This review was based on a search of the online scientific databases ( Scientific Information Database ) PubMed and Google Scholar from 1952 through 31st August 2016 . Keywords—submitted in English , French , Turkish and Russian—for the search were “flea AND fauna AND Iran”; “Iran AND puce” , “Iran AND siphonaptera”; “Iran AND ectoparasite” . Searches were conducted in the titles , abstracts , keywords and full text . The majority of our knowledge on the Siphonaptera of Iran is derived from plague studies[23] , the concept of “telluric plague” is coeval with these researches[24] and studies of two flea specialists , the Iranian Farhang-Azad and the French J . M . Klein . In each case the flea species , its host , and location of sampling were extracted from the published papers . The flea distribution maps were prepared using ArcGIS ( ArcGIS version 9 . 3 , ESRI ) . An online software were used to further classify and quantify the shared and exclusive vertebrate hosts of fleas with the “family or subfamily” filtering criteria[25] .
The data for this study were extracted from about 100 relevant papers in English , French , Istanbul Turkish or Russian . Faunistic reviews of the medically significant fleas showed the presence of fleas through 31 Iranian provinces ( Fig 1 ) . In the old classification of Iran provinces used by Farhang-Azad ( 1972b ) , the Khorasan province , which was the largest province of Iran in the plague research era , is currently divided in three provinces namely Razavi Khorasan , North Khorasan , and South Khorasan . This means that the spatial scale of the flea range resolution is less accurate in the old literature as it covers a larger area where the flea and their host are not homogenously found . Based on the information in the studied papers , the sampling locations mainly were human houses , animal husbandry premises , poultry farms , and rodents’ burrows . According to the literature , about 117 species or subspecies of fleas belonging to 7 families and 35 genera have been described in Iran . Most flea species reported in the studied literature belonged to the Ceratophyllidae ( n = 33 ) , Leptopsyllidae ( n = 24 ) , Pulicidae ( n = 21 ) , Ctenophthalmidae ( n = 20 ) and Coptopsyllidae ( n = 9 ) families . The flea species of the Ischnopsyllidae ( bat-fleas ) and Vermipsyllidae ( carnivore-fleas ) families consisted of only 6 and 4 species of the whole collection respectively ( Tables 1 and 2 ) . The Ceratophyllidae , the more represented flea family , consisted of 33 species belonging to 6 genera , comprising Callopsylla , Ceratophyllus , Citellophilus , Myoxopsylla , Nosopsyllus and Paraceras . The Leptopsyllidae , bird and rodent fleas , consisted of 24 species consisting of 10 genera including Amphipsylla , Caenopsylla , Ctenophyllus , Frontopsylla , Leptopsylla , Mesopsylla , Ophthalmopsylla , Paradoxopsyllus , Peromyscopsylla and Phaenopsylla . The Ctenophthalmidae consisted of 20 species belonging to 7 genera comprising Ctenophthalmus , Doratopsylla , Neopsylla , Palaeopsylla , Rhadinopsylla , Stenoponia and Wagnerina . The Pulicidae , a cosmopolitan family of the most notorious plague vectors ( genus Xenopsylla ) , included 21 species distributed in 7 genera comprising Archaeopsylla , Ctenocephalides , Echidnophaga , Pulex , Synosternus , Parapulex , and Xenopsylla . The Coptopsyllidae was limited to 9 species in the genus Coptopsylla . In the above-mentioned five families , the most commonly reported fleas belong to the genera Nosopsyllus ( Ceratophyllinae ) , Xenopsylla ( Xenopsyllinae ) , Ctenophthalmus ( Ctenophthalminae ) Coptopsylla ( Coptopsyllidae ) Amphipsylla ( Amphipsyllinae ) , Leptopsylla ( Leptopsyllinae ) , and Mesopsylla ( Mesopsyllinae ) . Detailed information is presented in Table 1 . The hosts associated with Iran fleas ranged from the small mammals ( Rodentia , Chiroptera , Lagomorpha , Insectivora ) to the large mammals ( Ungulata , Carnivora , Primates , Artiodactyla ) and birds as well . On the whole , 166 vertebrate host species were reported infested by fleas in Iran in the literature including Pulicidae ( n = 56 ) , Ceratophyllidae ( n = 38 ) , Ctenophthalmidae ( n = 29 ) , Leptopsyllidae ( n = 22 ) , Coptopsyllidae ( n = 11 ) , Ischnopsyllidae ( n = 7 ) and Vermipsyllidae ( n = 3 ) . By filtering the compiled data , we recognized 77 vertebrate host species among all seven flea families . Eight potential mammals were hosted by ≤7 flea ( sub- ) family respectively; these were: Calomyscus bailwardi ( 7 ) , Meles meles ( 7 ) , Mus musculus ( 7 ) , Meriones vinogradovi ( 8 ) , Vulpes vulpes ( 8 ) , Cricetulus migratorius ( 9 ) , Meriones libycus ( 9 ) and Meriones persicus ( 11 ) . Actually flea ( sub- ) families can infest ≥10 vertebrate hosts were Xenopsyllinae ( n = 43 ) , Ceratophyllinae ( n = 37 ) , Archaeopsyllinae ( n = 20 ) , Ctenophthalminae ( n = 20 ) , Pulicinae ( n = 19 ) , Amphipsyllinae ( n = 17 ) , Stenoponiinae ( n = 12 ) and Coptopsyllidae ( n = 11 ) . Detailed information is presented in Table 3 . At least 23 , 6 , 5 , 5 and 1 host species are exclusively infested by Pulicidae , Ischnopsyllidae , Ceratophyllidae , Ctenophthalmidae and Leptopsyllidae respectively . However restricted host species was not found in the Coptopsyllidae and Vermipsyllidae ( Table 4 ) . A total of 53 vertebrate species were reported infested by six subfamilies of Ctenophthalmidae including Ctenophthalminae ( n = 20 ) , Stenoponiinae ( n = 12 ) , Rhadinopsyllinae ( n = 9 ) , Anomiopsyllinae ( n = 6 ) , Doratopsyllinae ( n = 3 ) and Neopsyllinae ( n = 3 ) . By filtering the compiled data , 29 vertebrate host species were distinguished among all six subfamilies . Correspondingly 8 , 6 and 1 host species are exclusively included in the Ctenophthalminae , Stenoponiinae and Doratopsyllinae . However there were not found any restricted vertebrate host species in the Anomiopsyllinae , Neopsyllinae and Rhadinopsyllinae ( Table 5 ) . A total of 33 vertebrate species were reported infested by three subfamilies of Leptopsyllidae including Amphipsyllinae ( n = 17 ) , Mesopsyllinae ( n = 9 ) and Leptopsyllinae ( n = 7 ) . By filtering the compiled data , 22 vertebrate host species were distinguished among three subfamilies . Investigation on the flea-host associations in subfamilies of the Leptopsyllidae showed that there were no common host species shared by the three subfamilies . However 6 , 3 and 2 host species are exclusively included in the Amphipsyllinae , Leptopsyllinae and Mesopsyllinae respectively ( Table 6 ) . A total of 83 vertebrate species were reported infested by three subfamilies of Pulicidae including Xenopsyllinae ( n = 43 ) , Pulicinae ( n = 20 ) and Archaeopsyllinae ( n = 20 ) . By filtering the compiled data , 56 vertebrate host species were distinguished among three subfamilies . Exploration of flea-host associations in Pulicidae pointed out that there are eight common hosts including Capra hircus ( Linnaeus , 1758 ) , Hemiechinus auritus ( Gmelin , 1770 ) , Herpestes auropunctatus ( Hodgson , 1836 ) , Hyaena hyaena ( Linnaeus , 1758 ) , Meles meles ( Linnaeus , 1758 ) , Ovis aries ( Linnaeus , 1758 ) , Rattus rattus ( Linnaeus , 1758 ) and Vulpes vulpes ( Linnaeus , 1758 ) among three subfamilies . Although a number of 27 , 5 and 5 host species are exclusively included in the Xenopsyllinae , Pulicinae and Archaeopsyllinae respectively ( Table 7 ) .
The first step in identifying the risk linked to flea exposure is to make a list of the species before any public health measures can be taken . Flea-borne diseases are caused by emerging and re-emerging infectious agents which distribution , prevalence and incidence are currently increasing . However , the data about fleas and their medical significance in different geographical regions of Iran is limited . We took the first step in this paper but supplementary studies are required to i ) complete the list , especially in areas where there are no reportsor poor faunistic studies and ii ) perform molecular screening of flea pools in order to detect specific pathogen circulation in domestic fauna and wildlife in order to prevent future epidemics . | The data about flea-borne emerging or re-emerging infections throughout Iran are limited . This paper showed that the flea fauna of Iran were dominated by seven families . Moreover flea-host association analysis indicates that rodents are common hosts of flea families and most vertebrates hosted fleas belonging to the subfamilies Xenopsyllinae , Ctenophthalminae and Amphipsyllinae . We showed that the Persian jird ( Merions persicus Blanford , 1875 ) and the Xenopsyllinae are respectively the major vertebrate and potential vectors of flea-borne diseases in Iran . Further efforts are needed to inventorize and screen molecularly wild and domestic mammals flea fauna ( >3kg ) in order to monitor the risk of and control flea-borne infections in Iran , especially in the ecoregions with high diversity of flea and host species and in the old endemic plague foci of the country . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"plagues",
"ecology",
"and",
"environmental",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"geographical",
"locations",
"microbiology",
"vertebrates",
"animals",
"bacterial",
"diseases",
"yersinia",
"habitats",
"fleas",
"bacteria",
"bacterial",
"pathogens",
"infectious",
"diseases",
"yersinia",
"pestis",
"medical",
"microbiology",
"microbial",
"pathogens",
"pathogenesis",
"insects",
"iran",
"arthropoda",
"people",
"and",
"places",
"asia",
"host-pathogen",
"interactions",
"biology",
"and",
"life",
"sciences",
"organisms"
] | 2017 | The Fleas (Siphonaptera) in Iran: Diversity, Host Range, and Medical Importance |
The ability to integrate experiential information and recall it in the form of memory is observed in a wide range of taxa , and is a hallmark of highly derived nervous systems . Storage of past experiences is critical for adaptive behaviors that anticipate both adverse and positive environmental factors . The process of memory formation and consolidation involve many synchronized biological events including gene transcription , protein modification , and intracellular trafficking: However , many of these molecular mechanisms remain illusive . With Drosophila as a model system we use a nonassociative memory paradigm and a systems level approach to uncover novel transcriptional patterns . RNA sequencing of Drosophila heads during and after memory formation identified a number of novel memory genes . Tracking the dynamic expression of these genes over time revealed complex gene networks involved in long term memory . In particular , this study focuses on two functional gene clusters of signal peptides and proteases . Bioinformatics network analysis and prediction in combination with high-throughput RNA sequencing identified previously unknown memory genes , which when genetically knocked down resulted in behaviorally validated memory defects .
The ability to form a memory is one of the hallmarks of the advanced nervous system . This capacity to learn from , and remember , past experiences is a critical attribute to many eukaryotes . It is this function that allows organisms to meet the various demands and challenges of a changing and stochastic world: The interruption of these processes is associated with devastating illnesses , such as Alzheimer’s Disease and Huntington’s Disease , amongst others . Given the importance of this facility , it is vital for us to understand the basic biological mechanisms behind the complex process of learning and memory . A traditional view of learning and memory involves two phases , acquisition and retention . However , decades of investigation have revealed this to be an oversimplification , and instead , a variety of studies point to the existence of distinct features for learning , short term memory ( STM ) , long term memory ( LTM ) , memory consolidation , and memory retrieval [1–4] . Differences in mechanism and neural circuitry for these functions point to a degree of autonomy for these related processes [5 , 6] . For instance , it has been suggested that disruption in memory retrieval may be a pathology distinct from other memory impairments [7] . This notion is further supported by the identification of discrete waves of transcriptional activity and protein synthesis associating spatially and temporally with the various learning and memory processes [8 , 9] . One convention in the field is the distinction between short term and long term memory . Intrinsic to the process of STM is the transient nature of the memory . Mechanistically it relies predominantly on reversible processes , such as protein modifications that alter synaptic function [3] . Altered function of the synapse is not persistent in STM , lasting from minutes to several hours . Additionally , it is functionally unique in that STM can respond rapidly to stimulus and is generally characterized by being independent of de novo protein synthesis [10] . Alternatively , LTM relies on the time and energy consuming processes of transcription and translation . Formation of LTM creates persistent and stable alterations of the synapse including both changes in synaptic connections , as well as synaptic potentiation . These changes can last for hours to weeks and are a combined result of transcription factor activation , protein synthesis , and synaptic protein reorganization [11 , 12] . Dynamic gene transcription is central to synaptic plasticity , with distinct waves of gene expression defining the various phases of LTM formation [2 , 13 , 14] . Immediate early gene ( IEG ) transcription occurs rapidly in response to neuronal activity . Within minutes , key IEGs are up regulate . This rapid response to stimulus is made possible by chromatin accessibility and de novo protein synthesis independence [15 , 16] . In mammalian systems IEGs are enriched for transcription factors and are required for triggering later transcriptional waves [17 , 18] . In Drosophila , these activity-regulated genes are less well characterized; although studies indicate that the general dynamics of expression persist these ARGs appear to include a more functionally diverse set of genes [19] . Later waves of gene transcription encode key memory genes responsible for neural plasticity and memory establishment [20] . In general , the specific activity-related and subsequent target genes that change in expression seem to be paradigm specific [18 , 21] . In addition , these subsequent transcriptional waves are classified as either cycloheximide sensitive or cycloheximide insensitive , pointing to multiple mechanisms of transcriptional regulation that are distinguished by requiring new protein synthesis [22 , 23] . The identification of individual memory genes , such as Fragile-X mental retardation gene 1 ( FMR1 ) and rutabaga ( adenylate cyclase ) , has enhanced our knowledge of specific molecular processes in the neurons [24 , 25] . However , genes are likely acting in complex networks rather than having a singular effect on neuron function . In this regard , much remains unclear and a global picture of gene expression networks has not yet been established . In part , progress has been hindered by the technical challenges associated with traditional learning and memory paradigms . In particular , the high degree of variability between individuals and tissue types has created roadblocks to the systems level analysis common in other fields . Classical studies involving associative learning paradigms and courtship rejection have identified critical learning and memory circuits in the Drosophila brain . The mushroom body ( MB ) , considered the learning and memory center in Drosophila , consolidates the multiple sensory inputs critical to learning and memory formation [1 , 26–28] . This study utilizes an alternative approach and recently developed Drosophila LTM paradigm involving the endoparasitoid wasp , Leptopilina heterotoma . These wasps are a natural predator of Drosophila larvae . Although adult flies are not at risk of parasitism , female flies have an innate response to the presence of these wasps , resulting in a cascade of behavioral changes dependent on LTM formation [29 , 30] . Advantages to this paradigm include the robust response to the predator , and the persistence of this response across several days . We use this system to explore global changes in transcription over the course of memory formation . With a systems level approach of combining molecular , behavioral and computational methods allowed us to discover several previously unidentified LTM genes . In addition , by using a bioinformatics methodology we begin to place these genes in a larger context that distinguishes LTM formation from LTM maintenance .
It has been previously observed that when Drosophila melanogaster are cohabitated with the parasitic wasp Leptopilina heterotoma ( LH14 ) , flies will seek out ethanol containing food [29 , 30] . Further , when wasps are removed from the environment , female flies continue to favor ethanol-containing food as an oviposition substrate , a behavior that persists through the process of long-term memory formation . Using a similar experimental design we were able to replicate these previous findings ( Fig 1A ) . Eggs were counted at the end of a 24-hour period , the proportion of eggs laid on ethanol food was calculated for each cage , giving an ethanol preference index; where a proportion of 0 . 5 , or 50% , would indicate indifference to the presence of ethanol food . Under our environmental conditions , the baseline of ethanol preference in unexposed flies was approximately 20–30% ( Fig 1D and 1E ) , signifying an avoidance of ethanol food . However , in the presence of wasps this ethanol preference increases to 93% ( p-value 1 . 08E-5 ) in wild type Canton S ( CS ) flies ( Fig 1D ) , illustrating a strong affinity for ethanol food under these conditions . In wild type flies , the ethanol preference is maintained following wasp exposure period , with female flies displaying an ethanol preference of 94% in the absence of wasps , which is significantly increased when compared to 32% unexposed ( p-value 1 . 8E-4 ) ( Fig 1E ) . These findings illustrate a behavioral switch perpetuated by the memory formed during wasp exposure . This ethanol seeking behavior has been shown to be robust across genetic backgrounds , and the maintenance of the ethanol seeking behavior is reliant on long-term memory formation . As expected , when we tested the classic memory mutants dnc1 and Orb2ΔQ , flies were able to respond to wasps , increasing from 19% and 19% to 96% and 94% ethanol preference respectively ( p-values 1 . 5E-4 , 1 . 7E-4 ) . However , these mutants were not able to form memory , showing no significant increase in ethanol-seeking behavior after the wasp exposure period ( p-values 0 . 96 dnc1 , 0 . 42 Orb2ΔQ ) ( Fig 1D and 1E ) . These data support the previous finding that ethanol seeking post-wasp exposure is dependent on long-term memory formation . To explore gene expression changes that correspond to this long-term memory formation , we sequenced four-day wasp exposed and paired unexposed head samples . A total of 165 genes had at least a log2 fold increase of 2 , and 14 genes were decreased by log2 fold of 2 or greater , with a false discovery rate ( FDR ) of 0 . 05 or less ( Fig 2A , S1 Table ) . Of these differentially expressed genes , six functional gene clusters were identified as enriched in a DAVID analysis with enrichment greater than three . In order from highest to lowest enrichment clusters are as follows; chitin binding and extracellular region , signal peptides , attacin-related , proteases , glycosidase and sucrose metabolism , and defense response to fungus ( S1 Fig , S2 Table ) . Similar results were observed when the fold change restriction was removed , and all genes with significant FDR were used as input . This analysis identified five functional clusters with enrichment greater than three . In descending order , these clusters are annotated as chitin binding , signal peptides , DM9 domain proteins , sugar metabolism , and proteases ( S3 Table ) . The DAVID analysis generated from the gene list with both log2 fold change cutoff and significant FDR returned two groups of particular interest: A protease cluster with enrichment of 4 . 9 , and a signal peptide cluster with enrichment of 8 . 1 ( Fig 2B , S2 & S3 Figs ) . Given the potential biological relevance of these groups , genes from these clusters with a minimum log2 fold change of 2 were used to generate interaction networks through Integrative Multi-species Prediction ( IMP ) . IMP predicts and graphically displays interaction between the genes as well as key genes predicted to be part of the network ( Fig 2C and 2D ) . It is noteworthy that the trypsin genes are cross-listed in both protease and signal peptide clusters , and this fact is partly responsible for some overlap between the two interaction networks . It is reasonable to assume that these clusters are integrated into larger networks and pathways , and possibly interact with one another . One observation in support of this possibility is the presence of Jon65Aiii , a member of the protease cluster , as a predicted node of the signal peptide network . Additional genes with overlap between the clusters are the predicted node of Tsp2A as well as κTry . Of interest , κTry is the gene with the greatest number of interactions in both clusters totaling 13 edges in signal peptide network , and 18 edges in protease network . Overall , the protease cluster was more highly connected and this necessitated the use of more stringent prediction parameters for this cluster to focus the analysis . In this network , after κTry the genes with the greatest number of interactions are CG10911 ( 17 edges ) , CG3868 ( 15 edges ) , and αTry ( 14 edges ) ( Fig 2D ) . The most connected genes in the signal peptide cluster are Amy-d , βTry , Tsp2A with 12 edges each ( Fig 2C ) . A subset of genes identified by sequencing was further confirmed by quantitative polymerase chain reaction ( qPCR ) and validated as significantly differentially expressed ( Fig 2E , S4 Table ) . PGRP-sb2 was used as a negative control , as it was not identified as differentially regulated in the sequencing data . P-values for all tests can be found in S5 Table . Classical memory genes were noticeably absent for the list of differentially expressed genes; many classical memory genes , such as rut , dnc , and amn , were identified through mutagenesis screens rather than differential gene expression [31–33] . In addition , more recent memory studies have not observed these traditional memory genes to be differentially expressed during memory formation on a global scale [34 , 35] . Therefore , the absence of these genes in our differentially expressed gene list is not surprising . However , two previously identified memory genes were present in the sequencing list with significant FDR: RYa-R , a neuropeptide receptor ( log2 fold change 3 . 4 , FDR 1 . 25e-7 ) and scb , a member of the integrin alpha chain family , ( log2 fold change 0 . 57 , FDR 1 . 37e-4 ) [36 , 37] . These findings function as a partial positive control within the data set . The process of memory formation is not instantaneous . In most learning paradigms , the animal requires repeated training sessions or extended duration of exposure to the stimulus for memory formation to occur . To capture transcriptional changes involved in the early stages of memory consolidation , we titrated the length of exposure to identify a wasp-exposure interval that did not confer the ethanol seeking behavior . Time points of 2 . 5 , 7 , and 14 hours were tested . We observed that neither 2 . 5 hours nor 7 hours of exposure is sufficient to trigger a behavioral switch in these flies . However , by 14 hours memory has been formed , as seen by the ethanol preference of 95% in the exposed group compared to 18% in the unexposed group ( p-value 1 . 6E-4 ) ( Fig 3A ) . Given these data , we hypothesized that unique memory-related genes not detected in the 4-day wasp-exposed samples , may be differentially regulated at these earlier time points . To explore this possibility , sequencing data was generated from female fly heads with 2 . 5 and 7 hours of exposure ( Fig 3B and 3C ) . The 2 . 5-hour time point revealed minimal differential gene expression ( Fig 3B , Fig 4A and 4B ) . Two genes , Bsg25A and Elba3 , were identified as having both a log2 fold change magnitude of 2 and significant FDR ( FDR = <0 . 05 ) . No genes with significant FDR were identified as overlapping with the 4-day differentially expressed gene list ( Fig 4A ) . At the 7-hour time point , 1693 genes were differentially expressed with a significant FDR , 93 of which had a minimum fold change magnitude of 4 . Down regulated genes constituted the bulk of this gene list , with 79 genes down regulated compared to only 14 genes up regulated ( S6 Table ) . When compared to the 4-day sequencing data , 10 genes were found to overlap between the two gene sets . Of these shared genes , only two genes are differentially regulated in the same direction; Jon65Aiii and Jon65Aiv are both up regulated ( Fig 4C , Fig 5A , S7 Table ) . Interestingly , Bsg25A and Elba3 are also down regulated at the 7-hour time point , resulting in complete overlap of the 2 . 5-hour gene list ( Fig 4B ) . We further examined the temporal dynamics of gene expression by measuring RNA levels following a 14-hour wasp exposure ( Fig 5 , S4 Table ) . Considerable overlap exists between the 4-day and 14-hour time point: For instance , αTry , βTry , yip7 , Jon65AiV , and Jon65Aiii were up-regulated at 14 hours , but to a lesser degree than the 4-day time point . Alternatively , IM18 and CecA1 had similar expression levels at the two time points . These differences in gene expression patterns hint at multiple regulatory pathways governing gene expression . Certain genes may reach their maximum induction quickly , or have their mRNA strictly regulated , resulting in plateaued RNA levels between the 14-hour and 4-day time points . Other genes may maintain increasing mRNA production as there is continuing memory formation or prolonged exposure to stress conditions . In addition to dynamics of a specific gene across time , gene-gene interactions may be key to the memory formation process as well . Although the gene sets show limited overlap in differentially expressed genes , it is noteworthy that signal peptides were found to be enriched in a DAVID analysis of the 7-hour data set ( S8 Table ) . Many of these genes are different from those in the 4-day signal peptide cluster; however , IMP network prediction suggests possible interactions between genes in the two clusters ( Fig 5B ) . Further , the network analysis predicted nodes within the 4-day signal peptide cluster that are differentially regulated in the 7-hour samples . In other words , genes that are predicted to interact with the 4-day signal peptide cluster are differentially regulated at an earlier time . Specifically , CecA1 , Dpt and Dro are up regulated in the 4-day exposed samples; these genes are known to interact with CecA2 , DptB , and AttA ( Fig 2C ) . These three predicted nodes of CecA2 , DptB , and AttA are differentially regulated with significant FDR , although only CecA2 meets the log2 fold change threshold; with log2 fold change values of -2 . 86 , -1 . 3 , and -1 . 85 respectively . To explore the role of these genes following memory formation , we measured transcript abundance in flies after a 24-hour recovery period following wasp removal . Although several genes exhibited a trend towards up or down regulation , none of the genes assayed had statistically significant differential gene expression , likely due to the unusually high variance in these samples ( Fig 5A , S4 Table ) . To evaluate the functional significance of the differentially expressed genes and their possible role in memory formation , we conducted behavioral experiments paired with gene knock down . Initial experiments were preformed with the Elav-Gal4 switch driver , an inducible pan neuronal driver , with the specific advantage of the Gal4 transcription factor being active only when the RU486 ligand is present . In this system the RU486 must be externally provided , thus allowing for temporal control of the RNAi expression ( Fig 6A and 6B ) . Elav-Gal4 switch lines were crossed to UAS-RNAi lines for a subset of differentially regulated genes; βTry , Dpt , Kaz-M1 , Jon65Aiii , IM18 , yip7 , Jon65Aiv , αTry , MalA1 . κTry and Tsp2A were additionally tested as both were predicted to be an interacting gene in the IMP analysis although not differentially expressed . Expression level of κTry was confirmed with qPCR as not differentially expressed following wasp exposure ( S4 Table ) . All of these genotypes had significant ethanol preference in the acute assays for vehicle only as well as with RU486 feeding , indicating that vehicle or RNAi depletion of each gene did not cause any defects in animals' ability to perceive and respond to the presence of the wasp predator ( Fig 6C and 6D ) . All 10 genotypes had memory formation following wasp exposure with vehicle only feeding , indicating that ingestion of vehicle ( 5% methanol ) did not disrupt memory formation ( Fig 6E ) . However , after wasp removal Kaz-M1 , Jon65Aiii , IM18 , yip7 , αTry , MalA1 , κTry and Tsp2A displayed memory defects upon RNAi knock down ( Fig 6F ) . These data indicate that each genotype is able to respond to wasps and form memory when the RNA-hairpin is not expressed . In addition , these results show that the RNA-hairpin expression does not inhibit the ethanol-seeking wasp response in the presence of wasps , but rather after wasp removal , RNAi depletion of certain genes interrupts memory formation or maintenance . Several attractive candidate genes did not result in memory defects , and we offer two conceivable explanations beyond the hypothesis that they are not relevant memory genes . First , it is possible that we had insufficient knock down of the gene to illicit a phenotype . Secondly , it is also reasonable to consider that , as the sequencing was performed on whole heads , some of the differentially expressed genes may be in non-neuronal tissues . In such a case our pan-neuronal driver would not express the RNA-hairpin in the relevant cell type . To understand where in the nervous system these genes function , we used a more specific Gal4 driver line , which expresses in the learning and memory center of the Drosophila brain , known as the mushroom body ( MB ) . Genes that yielded memory defects in the previous experiments were tested with this more specific MB-Gal4 switch driver . All genotypes tested had ethanol seeking behavior in the presence of wasps ( Fig 7A and 7B ) . Additionally , memory formation of the genotypes was not disabled when treated with vehicle only ( Fig 7C ) . However , IM18 , Jon65Aiii , αTry , and ĸTry knock down in the MB caused memory defects . We considered the possibility that RNAi depletion of any essential gene in MB neurons could damage or kill cells important for memory consolidation or maintenance . In this scenario such genes would not be essential for memory per se , but instead they could simply be required for cell survival or neuronal activity . To test this possibility flies were treated with RU486 to induce RNAi depletion , as preformed previously , and then allowed to ‘recover’ for 4 days without RU486 feeding . Subsequently , they were tested for memory formation after exposure to predatory wasp . We found that RNAi knock down of IM18 and alpha Try before wasp exposure did not lead to disruption of memory ( Fig 8 ) . These observations suggest that knock down of these genes does not permanently damage the neurons of the mushroom body , which would prevent memory formation as an indirect consequence of gross neurological defect . Therefore , we conclude that IM18 and αTry are likely to be required for memory formation or maintenance as opposed to indirectly disrupting memory by affecting general neuronal processes and survival . Of the validated genes resulting in memory defects in the signal peptide annotation cluster , little is known about functional processes , making speculation about their mechanistic role in the brain challenging . For instance , IM18 has no identified protein similarity nor does it have a described function [38] . Kaz-m1 has been studied in greater detail , located in the E ( spl ) -C locus it is a Kazal family protease inhibitor [39]: Yet , basic information on binding partners and localization remain unknown , making specific predictions impractical . However , general trends may point to larger scale pathways and processes . Jak/Stat signaling has been observed in the Kenyon cells of the mushroom body , leading to the hypotheses that immune signaling is triggering actin cytoskeleton arrangement or chromatin remodeling as part of memory formation [40] . Alternatively , it is possible that the activation of immune pathways is indicative of synaptic pruning similar to what is observed during nervous system development [41] . The protease annotation cluster has few well-characterized genes , but predicted functions of these genes and expression patterns provide useful information nonetheless . Yip7 RNA expression has been observed in Drosophila surface glia , although its role in these cells remains unknown [42] . More generally , serine proteases are suggested to interact with protease activated receptors . Further , Trypsin-like proteins have been observed at presynaptic terminals in mammalian models , leading to speculation that such genes are involved in synaptic plasticity and long term potentiation [43 , 44] . It is possible that similar processes are occurring in the Drosophila nervous system and that the up regulation of these genes during memory formation is part of synaptic remodeling .
Memory formation , maintenance , and retrieval occur through an intricate system where information from new sensory inputs and existing neurocircuitry is consolidated . This process is massively complex and to date we lack a clear , global understanding of it . In the case of LTM , memory formation relies on basic functions , such as gene expression and protein synthesis . Dissecting these mechanisms and their dynamics brings the field closer to this large-scale view of learning and memory . In this study we approached the question of gene expression dynamics during memory formation in a novel non-associative LTM paradigm . We confirm previous findings that wasp exposure triggers a LTM dependent behavioral change resulting in female flies preferring ethanol-containing food as an oviposition substrate; and report RNA sequencing data specific to this wasp induced LTM . It is well established that dynamic gene expression is necessary for persistent memory [2] . In this sense the results presented in this paper may not be surprising , where we identified 179 genes that were differentially regulated following a four day exposure to wasps . It is noteworthy , however , that more than 90% of these genes were up regulated . Previous gene expression studies have shown conflicting results in this regard . The distribution of differentially expressed genes is varied and seems to depend on the neuron type , paradigm , and timing of collection [18 , 21 , 45] . Given that samples were collected immediately following the removal of the wasps , it is possible that a number of these genes are activity regulated genes ( ARGs ) , which are typically up regulated following neuronal firing [16 , 17]; although our sequencing data from earlier time points does not provide strong support for this hypothesis . Alternatively , this observation may be a behavioral paradigm specific phenomenon , possibly unique to non-associative learning . Based on this data we identified six enriched functional gene clusters from our DAVID analysis . The gene cluster with highest enrichment related to chitin binding . A previous study has implicated peptidoglycans in the behavioral changes resulting from bacterial infection [46] . Enrichment of genes with these functional annotations may hint at a similar role for them in the modification of other defense related behaviors . In addition to this , two functional clusters , signal peptides and proteases , were of particular interest based on biological inference . These experiments have therefore generated a substantial candidate list of novel up regulated genes , some of which may be important for memory . The signal peptide cluster was of particular interest , as it contained a number of immune associated genes . In particular , immune deficiency pathway ( IMD ) components appeared in the sequencing , in addition to two immune inducible genes IM18 and AttB . The IMD genes identified included CecA1 , CecA2 , Dro , AttA , Dpt , and DptB and were differentially regulated at various time points . These findings raise questions about the role of immune genes in learning and memory . The immune system in Drosophila has long been linked to inflammation and neurodegeneration [47–50]: Yet this would be a surprising discovery in a LTM paradigm . We suggest that a more likely scenario relates to recent studies that have been revealing a larger role for immune peptides in the nervous system , such as their participation in neuronal differentiation [51–53] . It is becoming clear that immune genes , and the IMD pathway , have non-canonical functions in the nervous system . Sleep regulation is one key area in which these genes are being examined . In particular Dro and AttB increase in expression with sleep deprivation , and more generally IMD genes are involved in sleep regulation [54 , 55] . Given the profound effect of sleep on learning and memory , we speculate that these immune genes are contributing to neuronal function in some way . Another tantalizing observation is that a STAT92E isoform is up regulated following courtship rejection training , perhaps implicating immune regulatory networks outside of the IMD pathway [35] . The precise mechanism underlying the importance of immune genes in the brain currently remains unclear . However , it has been speculated that perhaps immune signaling pathways are used as communication between neuron and non-neuronal tissues , such as the fat body [55 , 56] . Additional hypotheses have been put forth that focus on post-transcriptional actions , for instance NF-ĸB , the upstream activator of several immune pathways , has been implicated in affecting receptor density and synaptic stability [57 , 58] . Although this and other papers have presented plausible evidence for the role of immune genes in neurons; it is nonetheless important to consider indirect , system wide effects of the immune system and related processes . It has been shown that long term memory formation is energy demanding , as it requires protein synthesis and synaptic remodeling . These energy demands have measurable phenotypic outcomes , for instance one study found that Drosophila with enhanced memory in turn had reduced resiliency to starvation and dehydration [59] . Conversely , fruit flies under starvation conditions display impaired memory [60] . Such findings reinforce the notion that neurological processes such as long-term memory formation require tradeoffs . We would therefore be remiss to not consider the energy demands of the immune system . It is possible that these genes are negative regulators of the immune system , and by up-regulating these genes , energy is redirected from immunity to the brain . Alternatively , such genes may be protective against oxidative and immune related stress damage on the nervous system . Given the complexity of memory formation at the organism level , such hypotheses will need to be rigorously addressed in future works . The second gene cluster explored was the protease enrichment group . Two genes identified showed particularly interesting trend: Jon65Aiii and Jon65AiV were the only two genes up-regulated at both the 7-hour and four-day time points . Their increased gene expression at 7 hours , before memory formation has occurred may indicate a role for them in memory initiation or possibly unidentified ARGs . Of further note , Jon65Aiii is cross-listed between the predicted gene networks for the signal peptide and protease clusters . Although not experimentally verified , the overlap between networks illustrates the complex interactions that may be at play in LTM . The predicted gene nodes of κTry and Tsp2A , again shared between both networks , add additional emphasis to complex interactions between gene networks . Of the candidate genes , several were experimentally validated as LTM genes . Using a pan-neuronal conditional RNAi , we were able to show that Kaz-M1 , Jon65Aii , IM18 , yip7 , αTry , and MalA1 are essential genes for this memory paradigm . In addition , we used network prediction to identify genes possibly important to the memory formation process that are not differentially regulated in our data sets . This computational method produced κTry and Tsp2A as candidate memory genes; remarkably , RNAi knock down of these genes yielded a memory phenotype , further validating such predictive tools . It is important to emphasize that the induction of double-stranded RNA and RNAi depletion of any of these genes did not inhibit wasp perception or the behavioral response; instead the memory formation itself was interrupted . Additional nuances are being introduced to the field of study as the distinctions between LTM and memory consolidation are established [61] . The reinforcement and establishment of a memory outside of the traditional learning and memory centers of the brain , such as the hippocampus in mammalian systems , may be viewed as fundamentally different from LTM [62 , 63] . Inhibition of certain waves of protein synthesis appears to affect memory persistence , instead of memory formation , in some model systems [5 , 7] . However , the line separating these two processes is not well defined and the role of previously identified LTM genes may need to be evaluated in the memory consolidation process to truly understand their functional significance . Of the seven genes tested , IM18 , αTry , Jon65Aiii , and κTry showed memory defects , indicating that these genes have essential functions in the learning and memory center of the Drosophila brain . Identifying genes acting specifically within the MB was initially unexpected , as the expression data was generated from whole heads . It is possible that these genes have multiple functions within the brain , or possibly that they are highly expressed in enough cells to overcome this whole head dilution . Also surprising was the lack of differential gene expression at 2 . 5 hours . Previous data would suggest that ARGs should be activated within this time window . However , data also suggests that the shared number of ARGs across neuron type is limited , this may restrict the genes found due to the pooling effect caused by full head samples [19] . Future approaches that can detect single cell changes in gene expression will be particularly useful in spatially mapping how and where changes in gene regulation contribute to learning and memory . Overall , this study has identified novel genes involved in non-associative LTM , and we have attempted to place these genes into a larger network context . Three of the genes identified have critical roles in the MB , while the remaining memory genes are likely acting either up stream or down stream of the learning and memory center . Although the behavioral experimentation suggests important roles for these genes , their specific functions are not known and require additional experimentation to elucidate mechanism . These data also support the existing literature that point to the IMD pathway as an important player in learning and memory . With continued research and the use of bioinformatics tools we hope these data complement and inform future studies into the process of LTM formation , and in combination with a robust non-associative learning and memory approach , we propose that gene function can be further dissected into learning , memory consolidation , and memory maintenance activities .
Stocks were maintained on standard Drosophila cornmeal-molasses media at room temperature ( S9 Table ) . Memory formation experiments were conducted in vials ( 9 . 5 by 2 . 5 cm ) containing 40 female flies and 10 male flies . The wasp-exposed group had the addition of 20 female Lh14 wasps , as previously described [30] . Wasp exposures were maintained for 2 . 5 , 7 , 14 , and 96 hours ( 4 days ) . Memory formation was determined by ethanol preference , measured by a food choice assay . Briefly , one male and five female flies were placed into cages with two food sources , one with 0% ethanol and the other 6% ethanol food . Flies remain in the cages for 24 hours , at which point the food plates are collected and the number of eggs laid on each food source is counted in a blinded manner , such that the counter is not aware of genotype or treatment . As noted previously , the baseline ethanol preference/avoidance is sensitive to both temperature and humidity; all experiments reported were performed in a room with over head lighting and maintained at 25° with 30% humidity . All memory experiments measured ethanol-seeking behavior immediately following the removal of the wasps . Acute response experiments were conducted to determine the ability of the flies to respond to wasps irrespective of memory formation . These experiments were completed in similar fashion to the memory experiments , but with the addition of three female wasps in the cages of the exposed group during the food choice assay . Knock down experiments were performed using the UAS-Gal4 Switch system , where the Gal4 transcription factor becomes active only in the presence of RU486 [64 , 65] . Instant food impregnated with either the drug or vehicle only was used as the delivery system for this method of genetic manipulation . Two grams of instant food was hydrated with 8 mL of either RU486 ( 0 . 22 mg/mL ) in 5% methanol , or vehicle only ( 5% methanol ) . Flies were transferred to new food each day . Memory experiments consisted of four days of feeding concurrent with wasp exposure . Ethanol choice assays immediately followed the removal of wasp and RU486/methanol food . Alternatively , the acute response experiments were comprised of three days of feeding concurrent with wasp exposure , followed by the acute ethanol choice assay . In these acute response experiments the feeding protocol of drug or vehicle only was maintained during the ethanol choice assay . For the above-mentioned experiments , each is comprised of 10 cages per group unless otherwise noted . The data presented from these experiments is shown as a proportion of eggs laid on ethanol food compared to the total egg number from the cage and plotted as an average of the 10 replicates . Error bars were generated through bootstrapping the mean with 95% confidence . P-values were calculated from the Mann-Whitney U test . All statistics were calculated in R ( version 3 . 0 . 2 “Frisbee Sailing” ) , p-values for all tests can be found in S5 Table . Female flies were collected in 15 mL conical tubes , frozen in liquid nitrogen and briefly vortexed . Heads were separated using stackable steel sieves with pore size 710 , 425 , and 125 μm: Approximately 100 heads were collected for each replicate . Samples were maintained in Trizol at -80° until RNA isolation was performed using the miRNeasy Kit ( Qiagen ) with on-column DNase treatment . Four samples of each group were sent for sequencing on the illumina platform . Samples underwent rRNA reduction followed by random priming and were sequenced with a depth of 40 million reads . The adapters of short reads were trimmed by trimmomatic ( version 0 . 33 ) [66] . The short reads were then mapped to Drosophila melanogaster reference genome ( release 6 . 02 ) using STAR [67] . PCR duplicated short reads were removed by samtools ( version 0 . 1 . 19 ) [68] . Read counts per gene were calculated via bedtools [69] . The differentially expressed gene analysis was performed with edgeR [70 , 71] . Expression values for the four-day wasp exposure experiment were calculated by normalizing the exposed group to the paired control samples . Additional time course sequencing experiments for hours 0 , 2 . 5 , and 7; expression values were determined by comparing to the 0 hour time point samples . Heat maps were generated using hierarchical clustering and the R package pheatmap . The sequencing heat map included every gene that had a log2 fold change magnitude of 2 or more and significant FDR at any single time point . Genes with significant FDR ( FDR = < 0 . 05 ) and log2 fold change of 1 were used in the primary DAVID enrichment analysis [72 , 73] . Generation of the graphical illustration of the DAVID network required integration of the gene clusters and the corresponding fold changes from edgeR using a customized Perl script . The DAVID plot was subsequently created using a customized R script using package plotrix ( Version 3 . 6–1 ) and iGraph ( Version 1 . 0 . 1 ) with a Fruchterman and Reingold layout [74 , 75] . This code can be found at https://github . com/chenhao392/flyMemoryProject . Secondary DAVID enrichment analysis was conducted using all genes with significant FDR ( FDR = <0 . 05 ) regardless of the fold change value . Genes from these groups with at least log2 fold change of 2 were used to generate interaction networks in IMP ( http://imp . princeton . edu ) [76] . The clusters identified by DAVID were analyzed independently . The signal peptide cluster analysis used a 0 . 13 minimum prediction threshold and 10 additional gene nodes limit for the generation of the network . Given the increased number of interactions within the protease cluster , a more stringent threshold of 0 . 2 minimum prediction score and 8 additional gene nodes limit . The network analysis for both the 7-hour and 4-day signal peptide clusters used all 9 genes from the 7-hour time-point with at least a log2 fold change of 2 . The genes Amy-P and Yip7 from the 4-day cluster we used as input genes in the analysis; other genes from this cluster were identified by IMP as predicted interacting genes . cDNA was generated from RNA samples using the QuantiTect Reverse Transcription Kit ( Qiagen ) . The iTaq Universal SYBR Green ( BioRad ) was used for the PCR reaction , and all primers were validated with standard curve before use ( S10 Table ) . Gene specific data was normalized to actin and log2 fold change was calculated using the delta-delta CT method . An additional time point was collected post-memory formation: Following the 4 days of exposure , flies were separated from wasps and allowed to recover for 24 hours in a new vial before collection . All other time-points used RNA from the sequencing samples . Significance was determined by a two-tailed t-test . Statistical calculations were preformed in R ( version 3 . 0 . 2 “Frisbee Sailing” ) . | Long term memory formation is a complex process , and at different stages , requires regulation of transcription and protein synthesis . In a novel learning and memory paradigm , we examined transcriptional changes in the fly brain during and after memory formation . With RNA sequencing , we captured transcriptional waves of numerous genes previously not associated with learning and memory . Genes in the functional groups of proteases , signal peptides , and immunity , were selectively tested and behaviorally validated as memory genes . Placed into a large context and with computational methods , this work presents novel gene networks that may be linked to the learning and memory process . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genetic",
"networks",
"chemical",
"compounds",
"gene",
"regulation",
"enzymes",
"enzymology",
"neuroscience",
"organic",
"compounds",
"learning",
"and",
"memory",
"cognition",
"network",
"analysis",
"memory",
"alcohols",
"computer",
"and",
"information",
"sciences",
"proteins",
"gene",
"expression",
"chemistry",
"ethanol",
"biochemistry",
"organic",
"chemistry",
"post-translational",
"modification",
"gene",
"identification",
"and",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"proteases",
"physical",
"sciences",
"cognitive",
"science",
"signal",
"peptides"
] | 2017 | A systems level approach to temporal expression dynamics in Drosophila reveals clusters of long term memory genes |
It has long been known that multiple sclerosis ( MS ) is associated with an increased Epstein-Barr virus ( EBV ) seroprevalence and high immune reactivity to EBV and that infectious mononucleosis increases MS risk . This evidence led to postulate that EBV infection plays a role in MS etiopathogenesis , although the mechanisms are debated . This study was designed to assess the prevalence and magnitude of CD8+ T-cell responses to EBV latent ( EBNA-3A , LMP-2A ) and lytic ( BZLF-1 , BMLF-1 ) antigens in relapsing-remitting MS patients ( n = 113 ) and healthy donors ( HD ) ( n = 43 ) and to investigate whether the EBV-specific CD8+ T cell response correlates with disease activity , as defined by clinical evaluation and gadolinium-enhanced magnetic resonance imaging . Using HLA class I pentamers , lytic antigen-specific CD8+ T cell responses were detected in fewer untreated inactive MS patients than in active MS patients and HD while the frequency of CD8+ T cells specific for EBV lytic and latent antigens was higher in active and inactive MS patients , respectively . In contrast , the CD8+ T cell response to cytomegalovirus did not differ between HD and MS patients , irrespective of the disease phase . Marked differences in the prevalence of EBV-specific CD8+ T cell responses were observed in patients treated with interferon-β and natalizumab , two licensed drugs for relapsing-remitting MS . Longitudinal studies revealed expansion of CD8+ T cells specific for EBV lytic antigens during active disease in untreated MS patients but not in relapse-free , natalizumab-treated patients . Analysis of post-mortem MS brain samples showed expression of the EBV lytic protein BZLF-1 and interactions between cytotoxic CD8+ T cells and EBV lytically infected plasma cells in inflammatory white matter lesions and meninges . We therefore propose that inability to control EBV infection during inactive MS could set the stage for intracerebral viral reactivation and disease relapse .
Multiple sclerosis ( MS ) is the most common chronic inflammatory disease of the central nervous system ( CNS ) causing demyelination , neurodegeneration and disability . In most cases , MS is characterized by a relapsing-remitting course at onset which eventually develops into a progressive form; more rarely MS manifests as a primary progressive disease [1] . Immunomodulating and immunosuppressive drugs can reduce but not halt the disease process . Both the etiology and pathogenic mechanisms of MS are poorly understood . Genetic and environmental factors have been implicated in MS development , but the identity of the antigens ( self or non-self ) promoting chronic CNS inflammation remains elusive [2] . Several viruses have been linked to MS; however , Esptein-Barr virus ( EBV ) shows the strongest association with the disease [3]–[5] . EBV is a B-lymphotropic DNA herpesvirus that infects 95–98% of individuals worldwide , establishes a life-long , generally asymptomatic infection in B cells , and is the cause of infectious mononucleosis and of several lymphatic and non-lymphatic malignancies [6] . EBV has also been implicated in common autoimmune diseases , like systemic lupus erythematosus and rheumatoid arthritis [7] , [8] . Numerous studies have consistently demonstrated a higher prevalence of EBV infection and higher titers of antibodies to EBV antigens , in particular to EBV nuclear antigen-1 ( EBNA-1 ) , in young and adult MS patients compared to age-matched , healthy individuals [9]–[14] . It has also been shown that high titers of anti-EBNA-1 antibodies prior to MS onset [15] or at the time of a clinically isolated syndrome [16] and a previous history of infectious mononucleosis [17] increase the risk of developing MS . Furthermore , MS patients have higher frequencies of CD4+ T cells specific for EBNA-1 relatively to healthy , seropositive individuals [18] , while EBV-specific CD8+ T-cell responses in MS have been reported to be increased or decreased in different studies [19]–[23] . Although enhanced immune reactivity to EBV in MS suggests perturbed EBV infection , it is debated whether and how this can induce or influence the disease . EBV infection could contribute to MS through multiple mechanisms , including molecular mimicry , immortalization of autoantibody-producing B cell clones , and immunopathology [3] , [24] . It has been shown that CD4+ T cells from some MS patients cross-react with EBV and myelin antigens but the relevance of this finding to disease pathogenesis is still unclear [25] , [26] . EBV DNA load in the peripheral blood does not differ significantly between MS patients and healthy donors ( HD ) [16] , [18] , [23] and the possibility that a persistent EBV infection in the CNS drives an immunopathological response that damages myelin and neural cells is reasonable but remains controversial [27] . While several studies report absence of EBV in MS brain lesions [28]–[31] , we [32]–[34] and others [35] have shown that an abnormally high proportion of B cells infiltrating the MS brain are latently infected with EBV . We have also shown that ectopic B-cell follicles present in the inflamed meninges of patients with secondary progressive MS harbour EBV infected B cells and that EBV can reactivate in plasma cells in immunologically active white matter lesions and meningeal B-cell follicles [32]–[34] . If MS were the result of an immunopathological response aimed at eliminating a persistent EBV infection in the CNS , a positive correlation should be found between disease activity assessed by magnetic resonance imaging ( MRI ) or clinical progression and immune response to EBV . In support of this hypothesis , it has been shown that serum levels of EBNA-1 IgG positively correlate with gadolinium-enhancing MRI lesions ( characteristic of acute inflammation ) , lesion size and Expanded Disability Status Scale ( EDSS ) in patients with MS [36] and in patients with a clinically isolated syndrome who develop definite MS [16] . Another study reported higher disease activity on MRI in a subgroup of relapsing-remitting MS patients with stable levels of IgG specific for EBV early antigens expressed during the lytic cycle [37] . It has also been shown that CD8+ T-cell responses toward pooled EBV latent and lytic antigens in the blood of MS patients are high early in MS course and decrease during disease progression suggesting a possible association with more frequent episodes of CNS inflammation in early disease phases [21] . However , it is not known whether changes in the CD8+ T cell response to individual EBV latent and/or lytic antigens are associated with active and inactive MS phases . To address this issue , we have used pentamer staining to characterize the CD8+ T-cell response to EBV in the peripheral blood of patients with relapsing-remitting MS , both untreated and treated , and HLA-matched controls . Positivity to pentamer staining was then correlated with disease activity and inactivity , as assessed by clinical criteria and MRI of the brain . Our study reveals a lower prevalence of the CD8+ T cell response to EBV in inactive MS patients , a higher frequency of CD8+ T cells specific for EBV lytic and latent antigens during active and inactive disease , respectively , and marked changes in the EBV-specific CD8+ T cell response during treatment with approved disease-modifying drugs , such as interferon-β ( IFN-β ) and natalizumab . By analyzing post-mortem MS brain tissue , we demonstrate that the same EBV lytic antigen eliciting a higher CD8+ T cell response in the peripheral blood during active MS is expressed in inflammatory white matter lesions and meninges . We also show interactions between CNS-infiltrating cytotoxic CD8+ T cells and EBV lytically infected plasma cells , further supporting the link between EBV reactivation , higher cytotoxic immune responses to EBV lytic antigens and MS exacerbations .
Freshly isolated PBMC from HD ( n = 43 ) and untreated MS patients ( n = 79 ) were stained with the above mentioned fluorochrome-labeled HLA-A*0201 and/or HLA-B*0801/viral peptide epitope pentamers . We first examined the prevalence of EBV-specific CD8+ T cell responses in our cohort , namely the proportion of individuals with detectable pentamer+ CD8+ T cells ( Fig . S1 ) . Positive pentamer staining specific for at least one EBV epitope was found in a similar proportion of HD ( 39% ) and untreated MS patients ( 33% ) . The prevalence of CD8+ T cell responses to EBV latent and lytic antigens was similar in HD and untreated MS patients ( Figure 1 A ) . In HLA-A2+ subjects where both EBV- and CMV-specific CD8+ T cell responses could be evaluated , no differences in the response to either virus were found between HD ( n = 34 ) and untreated MS patients ( n = 45 ) ( Figure S2A ) . We then explored whether the prevalence of EBV-specific CD8+ T-cell responses could be related to disease activity . Untreated MS patients were subdivided in two groups consisting of 31 active and 48 inactive patients , as defined by presence and absence of clinical relapses and acute brain inflammation assessed with gadolinium-enhanced MRI , respectively ( Table 2 ) . There was a similar prevalence of latent and lytic antigen-specific CD8+ T cell responses in HD and active MS patients but a significantly lower prevalence of lytic antigen-specific CD8+ T cell responses in inactive MS patients than HD ( 19% versus 39% , p = 0 . 05 ) . Inactive MS patients also tended to have a lower prevalence of latent and lytic antigen-specific CD8+ T cell responses when compared with active MS patients ( 17% versus 35% , p = 0 . 1 and 19% versus 39% , p = 0 . 09 , respectively ) ( Figure 1A ) . In contrast , no differences were found among HD , inactive MS and active MS patients in the CD8+ T cell response to CMV ( Figure S2B ) . These findings therefore indicated a weaker immune response against EBV , particularly against lytic cycle antigens , in inactive MS patients compared to HD and active MS patients . Of note , 39% of active MS patients , but only 15% of inactive MS patients with a detectable EBV-specific CD8+ T cell response had a disease duration longer than 8 years , suggesting a decay with time of the EBV-specific immune response associated with inactive disease . We then evaluated the percentage of pentamer+ CD8+ T cells within the circulating CD3+ CD8+ T cell population in the study subjects with a detectable EBV-specific CD8+ T cell response . Similarly to prevalence , the frequency of latent and lytic antigen-specific CD8+ T cells did not differ significantly between HD ( n = 17 ) and total untreated MS patients ( n = 26 ) ( Figure 2 A ) . Also the frequency of CD8+ T cells specific for EBV and CMV antigens was similar in HLA-A2+ HD and total MS patients ( Figure S3A ) . Again , differences in the immune response to EBV emerged only when patients were stratified according to disease activity ( Figure 2 B–D ) . The frequency of latent antigen-specific CD8+ T cells in inactive MS patients ( 1 . 8±2 . 6% , mean ± SD ) tended to be higher than in HD ( 0 . 3±0 . 2% , mean ± SD; p = 0 . 07 ) and was significantly higher than in active MS patients ( 0 . 22±0 . 19% , mean ± SD; p = 0 . 05 ) ( cumulative data and representative plots are shown in Figure 2 B and D , respectively ) . This difference was mainly due to the fact that 6 out of 8 inactive MS patients recognized EBNA-3A and 3 of these displayed a very strong immune response against this EBV latent antigen ( 2 . 7 , 2 . 8 and 7 . 6% of total circulating CD8+ T cells were EBNA-3A-specific ) . In contrast , the frequency of lytic antigen-specific CD8+ T cells in active MS patients ( 1 . 8±2 . 8% , mean ± SD ) was significantly higher than in HD ( 0 . 34±0 . 28% , mean ± SD; p = 0 . 03 ) and tended to be higher than in inactive MS patients ( 0 . 3±0 . 2% , mean ± SD , p = 0 . 1 ) . The latter difference did not reach statistical significance probably due to the low number of inactive MS patients analyzed ( n = 9 ) . These findings therefore indicated more frequent recognition of EBV latent and lytic antigens during inactive and active MS phases , respectively . In contrast , the frequencies of CMV-specific CD8+ T cells did not differ significantly among HLA-A2+ HD , inactive MS and active MS patients ( Figure S3B , C ) . No correlation was found between the frequency of EBV-specific CD8+ T cells and disease duration in inactive and active MS patients ( Figure S4 ) . We then analyzed the EBV-specific CD8+ T cell response in MS patients who were treated with IFN-β ( n = 20 ) and natalizumab ( n = 14 ) ( Table 1 ) . IFN-β , the most frequently used first-line treatment for relapsing-remitting MS , has antiviral and immunoregulatory activities and reduces relapse frequency and brain MRI activity in relapsing-remitting MS patients [41] . The monoclonal antibody natalizumab inhibits lymphocyte extravasation into the CNS and is highly effective in suppressing clinical relapses and disease activity in relapsing-remitting MS patients who fail to respond to first-line therapies [42] . Among 20 MS patients treated with IFN-β for 1 to 11 years ( median = 4 years ) , 7 and 13 were in the active and inactive phase of disease , respectively ( Table 2 ) . The prevalence of the CD8+ T cell response to EBV latent and lytic antigens was similar in total untreated and IFN-β-treated MS patients ( Figure 1 A , B ) . However , none of the 13 IFN-β-treated inactive patients had a detectable CD8+ T cell response to EBV ( Figure 1 B ) and CD8+ T cells for CMV were found only in 1 of 9 HLA-A02+ IFN-β-treated inactive patients ( data not shown ) , indicating that effective IFN-β therapy is associated with a general inhibition of the antiviral response . In contrast , 57% ( 4/7 ) and 66% ( 4/6 ) of the IFN-β-treated MS patients with active disease had a detectable CD8+ T cell response to EBV ( Figure 1 B ) and CMV ( data not shown ) , respectively . Moreover , the frequency of CD8+ T cells specific for EBV latent , but not lytic , antigens was significantly higher in IFN-β-treated ( 0 . 8±0 . 6% , mean ± SD ) than in untreated ( 0 . 2±0 . 2% , mean ± SD ; p = 0 . 01 ) active MS patients ( Figure 2 B , C ) . After 8 to 16 months ( median = 12 months ) of treatment with natalizumab , all 14 MS patients analyzed were in the inactive phase of disease , both clinically and by MRI ( Table 2 ) . Unexpectedly , nearly all patients in this group had a detectable CD8+ T cell response to EBV latent and lytic antigens ( 87% and 93% , respectively ) ( Figure 1 B ) . This prevalence was significantly higher than that found in HD and any other group of MS patients ( p<0 . 001 ) . Nevertheless , the frequencies of latent and lytic antigen-specific CD8+ T cells in natalizumab-treated MS patients were similar to those found in untreated inactive MS patients ( Figure 2 B , C ) . As expected for patients in the inactive phase of disease , the frequency of lytic antigen-specific CD8+ T cells tended to be lower in natalizumab-treated patients than in untreated active MS patients ( p = 0 . 09 ) ( Figure 2 C ) . The frequency of CMV-specific CD8+ T cells did not differ among HD , untreated and natalizumab-treated MS patients ( data not shown ) . We then asked whether changes in the EBV-specific CD8+ T cell response during active and inactive MS phases could be detected in longitudinal studies . Despite experiencing clinical relapses , two patients ( HLA-B08+ B2/B2-2 and HLA-A02+ A14 ) in our cohort refused immunomodulatory therapy and agreed to be monitored periodically for 27 and 7 months , respectively . Patient B2/B2-2 , who was clinically silent and was diagnosed MS one year before our analysis started , displayed a very highy frequency of EBNA-3A-specific CD8+ T cells ( 6% of circulating CD8+ T cells ) at the beginning of the observation period ( month 0; Figure 3 A ) . EBNA-3A-specific CD8+ T cells progressively decreased during the subsequent months and became undetectable between month 17 and 27 . In parallel , the frequency of BZLF-1 specific CD8+ T cells , which were undetectable at previous time points , abruptly increased and reached a peak ( 11% of total circulating CD8+ T cells ) at month 21 in concomitance with the presence of active MRI lesions . The percentage of BZLF-1-specific CD8+ T cells in the CD8+ T cell population then declined to 2 . 5% in the subsequent 6 months , denoting marked expansion and subsequent contraction of the immune response toward this EBV lytic antigen ( Figure 3 A ) . In patient A14 , who experienced frequent clinical relapses , the frequency of CD8+ T cells specific for the EBV lytic antigen BMLF-1 ranged between 0 . 56 and 0 . 71% during a clinical relapse ( months 0 and 3 of the observation period ) and in the presence of an active MRI scan , and dropped to 0 . 11% in the subsequent 4 months ( Figure 3 A ) . During the same period , the frequency of CD8+ T cells specific for the EBV latent antigen LMP-2A and for CMV pp65 antigen remained low and stable ( Figure 3 A ) . Longitudinal analysis of EBV-specific CD8+ T-cell responses was also performed in 2 HLA-B08+ MS patients ( BTY5 and BTY8 ) treated with natalizumab and monitored periodically for 15 and 12 months , respectively , starting at 12–14 months after therapy initiation . Both natalizumab-treated patients were in the inactive phase of disease ( according to clinical and MRI evaluation ) and had a detectable CD8+ T cell response to EBNA-3A and BZLF-1 ( Figure 3 B ) . However , while the frequency of BZLF-1-specific CD8+ T cells was stable during the whole observation period a steady rise in the CD8+ T cell response to EBNA-3A was observed in both patients after 15–18 months of therapy ( Figure 3B ) . Three HD followed for 8 to 19 months did not show any significant variation in the frequency of CD8+ T cells specific for EBV latent and lytic antigens ( Figure 3 C ) . The increased frequency of CD8+ T cells specific for two EBV lytic antigens ( BZLF-1 , BMLF-1 ) in the blood of MS patients with an active MRI scan indirectly suggests a response to a previous or concomitant viral reactivation in the brain . In search for a link between immunological findings and brain inflammation , we examined the expression of BZLF-1 mRNA and protein in post-mortem brain tissue from 7 patients who died in the secondary progressive phase of MS and were characterized by a severe clinical course and substantial brain inflammation . The MS brain samples selected for this study contained immunologically active ( both active and chronic active ) lesions in the white matter and highly inflamed meninges with B-cell follicle-like structures that we previously showed to be major EBV reservoirs [32]–[34] . In preliminary experiments aimed at evaluating the specificity and binding of anti-BZLF-1 monoclonal antibody , we observed BZLF-1 immunoreactivity in the nucleus of EBV transformed B95-8 cells and in an EBV+ tonsil from a patient with infectious mononucleosis ( Figure S5 ) . Conversely , no BZLF-1+ cells were detected in sections of a non pathological human brain , of a brain from a patient with tuberculous meningoencephalitis and of an EBV-negative lymphoma ( Figure S5 ) . BZLF-1 immunoreactivity was detected in brain samples of all MS cases analyzed using immunohistochemical techniques ( n = 5 ) . Both isolated and small groups of BZLF-1+ cells were present in the inflamed meninges , at the periphery of B-cell follicle-like structures ( Figure 4 A–H ) and in diffuse inflammatory cell infiltrates ( Figure 4 I–K ) . BZLF-1+ cells were also present in the perivascular cuffs of inflamed blood vessels in active white matter lesions characterized by a high density of intraparenchymal foamy macrophages ( Figure 5 ) , but not in demyelinated , chronic active and inactive white matter lesions ( data not shown ) , thus linking EBV reactivation to acute inflammation . Nearly all BZLF-1+ cells infiltrating the meninges ( Figure 4 C , D , K ) and active WM lesions ( Figure 5 D , E , G , H ) were identified as Ig-producing plasmablasts/plasma cells , which is consistent with the knowledge that EBV reactivates upon B-cell differentiation into plasma cells [43] . At these sites the proportion of Ig+ cells expressing BZLF-1 ranged between 1 and 10% . BZLF-1+ plasma cells were detected in the same infiltrated brain areas where plasma cells expressing BFRF1 , an EBV early lytic protein induced by BZLF-1 [44] , were also found ( Figure 4 E , Figure 5 F ) . Expression of BZLF-1 was also investigated using quantitative real-time RT-PCR in 4 inflamed MS brain samples , 2 of which had been analyzed by immunohistochemistry . No BZLF-1 RNA was detected in whole MS brain sections ( data not shown ) . This negative result was expected given the paucity of EBV lytically infected plasma cells relatively to the large and heterogeneous cell population of the MS brain . To enrich for EBV infected cells and increase the sensitivity of the technique , perivascular and meningeal inflammatory cell infiltrates and the surrounding , non infiltrated brain parenchymal regions were harvested from MS brain sections using laser capture microdissection and analyzed using pre-amplification , quantitative real time RT-PCR for BZLF-1 . CD19 transcripts were also analyzed to optimally discriminate between infiltrated and non infiltrated brain areas . BZLF-1 transcripts were detected in the perivascular cuffs isolated from one active MS lesion and in 3 out of 4 meningeal B-cell follicles but not in 3 chronic active lesions , 4 sparse meningeal infiltrates and 7 non infiltrated parenchymal regions ( Figure 6 ) . A control lymph node was negative for BZLF-1 . Thus , both immunohistochemical and RT-PCR findings corroborated BZLF-1 expression and therefore shift to EBV lytic infection in immunologically active white matter lesions and ectopic B-cell follicles . Then , we searched for interactions between cytotoxic CD8+ T cells and EBV infected cells in the same MS brain samples in which BZLF-1 protein and/or RNA were detected . We first analyzed the presence and frequency of granzyme B-expressing CD8+ T cells and their relationship to EBV litically infected cells . We observed that most granzyme B+ cells in the MS brain co-expressed CD8 and that the fraction of CD8+ T cells expressing granzyme B ranged between 5 and 60% in the different MS cases and brain areas analyzed , the highest values being detected in the perivascular cuffs of active white matter lesions ( Figure 7 A , B ) . In the meninges , granzyme B+/CD8+ T cells were present in diffuse meningeal infiltrates and at the periphery of B-cell follicle-like structures , but were rarely seen inside these structures ( Figure 7 C , D ) . Given the nuclear localization of BZLF-1 and the relatively small number of BZLF-1+ cells in the MS brain it was extremely difficult to see contacts between granzyme B+ cells and lytically infected cells using double immunofluorescence for BZLF-1 and CD8 or granzyme B . We therefore stained MS brain sections for the EBV lytic protein BFRF1 which has a perinuclear localization and has been detected in a higher fraction of plasma cells ( up to 50% ) compared to BZLF-1 [32] . We observed lymphoblastoid CD8+ T cells adhering to or secreting granzyme B toward BFRF1+ cells as well as contacts between CD8+ T cells and BFRF1+ cells displaying granzyme B immunoreactivity on their surface ( Figure 7 E–H ) . Such cytotoxic contacts were frequently observed in sparse meningeal infiltrates and active white matter lesions , but not inside ectopic B-cell follicle-like structures . Finally , staining of MS brain sections for perforin and Ig allowed to visualize perforin granules polarized toward Ig+ cells inside the perivascular cuffs of active white matter lesions ( Figure 7 I , J ) , supporting further the idea that EBV harbouring cells might be the target of a cytotoxic attack .
Altered control of EBV infection in individuals susceptible to MS is suspected to play a role in the development of immune dysfunction causing CNS pathology [3]–[5] . Higher serum titers of EBNA-1 IgG are associated with an increased risk of MS [15] , increased conversion from a clinically isolated syndrome to definite MS [16] and more severe disease activity and clinical progression [16] , [36] . Virus-specific CD8+ T cell responses play an essential role in the control of EBV infection [45] and have been investigated in previous studies in MS using mainly IFN-γ ELISPOT analysis in PBMCs stimulated in vitro with EBV+ lymphoblastoid cells [20] , [22] , viral lysates [21] , individual [19] or pooled [21] EBV lytic and latent peptides , and more recently using MHC-peptide tetramer staining [23] . Several studies have shown that EBV-specific CD8+ T cell responses are significantly higher in MS than in HD or in patients with other inflammatory neurological diseases [19]–[21] . However , lack of significant differences between MS patients and controls [16] , [23] and even reduced frequency of EBV-specific CD8+ T cells in MS patients [22] have also been reported . Use of cryopreserved versus freshly isolated PBMCs , analysis of patients with relapsing-remitting and progressive MS courses , and lack of stratification of patients according to disease activity may have hampered a clear understanding of the possible link between EBV-specific CD8+ T cell responses and MS pathogenesis . To obtain an accurate pattern of CD8+ T cell in vivo specificities , we have used highly standardized flow cytometric analysis with EBV-specific pentamers , that unequivocally identify antigen-specific CD8+ T cells , on freshly isolated PBMCs obtained from HD and relapsing-remitting MS patients . Importantly , both untreated and treated MS patients were studied and disease activity was evaluated in most patients with gadolinium-enhanced MRI shortly before or at the time of blood collection . Such a rigorous selection justifies the relatively small number of MS patients analyzed . The first main finding of this study is that differences in the prevalence and magnitude of the CD8+ T cell response to certain EBV latent and lytic proteins between MS patients and HD and within the MS cohort emerge only when patients are stratified according to disease activity and inactivity . By showing that fewer inactive MS patients have a detectable CD8+ T cell response against EBV lytic antigens compared with HD and active MS patients and that the frequency of lytic antigen-specific CD8+ T cells is higher in active MS patients than in HD and inactive MS patients , we demonstrate for the first time that changes in the immune control of EBV replication are associated with the active and inactive phases of MS . This is corroborated by the longitudinal study performed in two untreated MS patients showing a peak in the frequency of CD8+ T cells to EBV lytic antigens during active disease . Of the two EBV lytic antigens analyzed , BZLF-1 is a transactivator expressed at the very initiation of the lytic cycle and is involved in the induction of early lytic proteins , including BMLF-1 [6] . Thus , an increase in BZLF-1- and BMLF-1-specific CD8+ T cells in concomitance with acute brain inflammation on MRI strongly suggests an attempt of the immune system to control intracerebral foci of EBV replication . On the other hand , a logical explanation for the decrease in lytic antigen-specific T cells associated with inactive MS could be elimination of lytically infected cells brought about by the strong cytotoxic response occurring in the active disease phase . In this context , it is important to recall that EBV DNA load in the blood of MS patients does not differ significantly from that in healthy EBV carriers [16] , [18] , [23] indicating that fluctuations in EBV-specific CD8+ T cell responses in MS patients do not result in a generalized impairment of the immune control of EBV infection . In contrast with the present findings , Jilek et al . [23] did not observe differences in the prevalence and frequency of CD8+ T cell responses to BZLF-1 and BMLF-1 between MS patients and control subjects . However , this study was not restricted to patients with relapsing-remitting MS , did not distinguish between patients with active and inactive disease and used cryopreserved PBMCs . Importantly , in our study both the prevalence and magnitude of the CD8+ T cell response to CMV were similar in HD and untreated MS patients , irrespective of disease activity , indicating that the differences observed in EBV-specific immunity are not the consequence of a general activation of antiviral immune responses driven by a still unknown MS-associated immune dysfunction . Despite the fact that the prevalence of the CD8+ T cell response to EBV latent antigens in inactive MS patients was similar to that in HD and tended to be lower than in active MS patients , we found that the magnitude of the CD8+ T cell response to EBNA-3A was higher in inactive MS patients than in HD and active MS patients . Very high numbers of EBNA-3A-specific CD8+ T cells were detected in half of the inactive MS patients harbouring this immune reactivity ( 2 . 7 to 7 . 6% of the circulating CD8+ T cells versus <1% in HD and active MS patients ) . Furthermore , longitudinal monitoring of a patient with a recent diagnosis of MS showed substantial reduction of EBNA-3A-specific CD8+ T cells just before the active disease phase and the rise of lytic antigen-specific CD8+ T cells . It is known that EBNA-3A is expressed shortly after EBV infection of B cells together with the whole set of EBV latent proteins ( EBNA-LP , −1 , −2 , 3A , 3B , and −3C , LMP-1 , LMP-2A , LMP-2B ) that are essential to drive infected B cells into proliferation ( latency III or growth program ) [6] and elicit strong T-cell responses [45] . Most EBV-encoded latent antigens , including EBNA-3A , are then sequentially downregulated as EBV establishes a persistent infection in memory B cells ( latency II and I programmes ) [6] . Thus , the study of EBNA-3A-specific CD8+ T cells suggests that at least in some inactive MS patients there is an attempt by CD8+ T cells to control abnormal expansion of a latently infected B-cell pool . A decrease in the immune response to EBNA-3A could reflect a change in EBV latency programme and , possibly , set the stage for switching to the lytic cycle . Of interest , more abundant EBNA-3A-specific CD8+ T cells were detected only in MS patients with a short disease duration ( <5 years ) . A decrease in immune reactivity toward EBNA-3A with disease progression could be due to reduced antigenic stimulation or to T-cell exhaustion which is known to occur during uncontrolled , chronic viral infections [46] . Relevant to this , we have shown that most EBV latently infected B cells accumulating in the inflamed MS brain during late-stage disease are memory B cells expressing the latency II programme [32] , [33] . Based on the present findings , we propose that failure to fully control EBV latent infection in an immune privileged site like the CNS could lead to recrudescence of EBV reactivation . Exposure to newly synthesized viral antigens would promote expansion of lytic antigen-specific CD8+ T-cells targeting intracerebral viral deposits and hence inducing the active phase of MS . Of relevance for the present findings , it has been shown that after primary EBV infection and during establishment of EBV persistence CD8+ T cells specific for some EBV epitopes disappear from the circulation after having upregulated Programmed Death-1 ( PD-1 ) inhibitory receptor , probably as a consequence of inadequate antigenic stimulation [47] . We are currently evaluating whether fluctuations in EBV-specific CD8+ T cells in relapsing-remitting MS might be associated with changes in PD-1 expression levels and T-cell function ( i . e . , cytokine profile and cytotoxic activity ) . It would be also interesting to compare the quality of the CD8+ T cell response to EBV in MS with that in systemic lupus erythematosus , an autoimmune disease associated with marked systemic EBV dysregulation [48] and impaired cytotoxic immune response to the virus [49] . The second main finding of this study is that treatment of relapsing-remitting MS patients with IFN-β and natalizumab is associated with marked changes in the CD8+ T cell response to viral antigens . We have shown that CD8+ T cells specific for EBV latent and lytic antigens and for CMV antigen were detectable in a substantial fraction of the patients entering active disease despite IFN-β treatment , but in none , except one , of the IFN-β-treated patients with inactive disease . It is likely that such a strong reduction in the CD8+ T cell response to both viruses might due to the direct antiviral activity of the drug [41] . Recently , Comabella et al . [50] reported that clinically effective IFN-β therapy is associated with downregulation of proliferative T cell responses to EBNA-1 without significant changes in the CD8+ T cell response against other ( pooled ) EBV antigens of the latent and lytic phase . Discrepancies with the present study could be due to technical issues , as discussed above . We have also shown that most natalizumab-treated MS patients , all of which were in the inactive phase of disease , had a detectable CD8+ T cell response to EBV . Such high prevalences could be related to the fact that natalizumab treatment causes a marked increase in lymphocyte numbers in the blood due to interference with lymphocyte extravasation and trafficking in lymphoid and non-lymphoid tissues [51] . Importantly , we found that in natalizumab-treated MS patients the frequency of CD8+ T cells specific for EBV lytic antigens was similar to that in HD and untreated inactive MS patients and stable over time ( up to 22–27 months of therapy ) . It therefore seems significant that the present analysis , though limited to a relatively small number of donors , consistently showed no expansion of EBV lytic antigen-specific CD8+ T cells during the inactive phase of MS regardless of presence or absence of therapy . In contrast , the frequency of EBNA-3A-specific CD8+ T cells , which was comparable in untreated and natalizumab-treated inactive MS patients within 8–16 months of therapy , progressively increased during the second year of therapy in 2 longitudinally monitored patients . These observations suggest dysregulation of EBV latent infection upon prolonged treatment with natalizumab . Although further studies are needed to clarify these aspects , analysis of EBV-specific CD8+ T cell responses in MS patients may help identify biomarkers useful for therapy monitoring and shed light into the mechanisms underlying drug efficacy . The third main finding of this study is that BZLF-1 , one of the two EBV lytic proteins recognized by CD8+ T cells expanding in the blood of active MS patients , is expressed in post-mortem MS brains with prominent immune infiltrates . The demonstration of BZLF-1 protein and RNA in active white matter lesions , which likely correspond to gadolinium-enhanced MRI lesions , and in the inflamed meninges , where changes in water content cannot be detected on MRI , lends support to the idea that acute brain inflammation in MS is associated with switch to the viral lytic cycle . In line with our previous results [32] , we have also shown that in the MS brain EBV reactivates in plasma cells and that the latter can be found in close contact with lymphoblastoid CD8+ T cells producing cytolotic enzymes . However , absence of CD8+ granzyme B+ T cells inside meningeal B-cell follicles , which contain a high frequency of EBV latently infected cells [32]–[34] , suggests that a local suppressive environment created by the virus itself to elude immune control [52] could hamper virus clearance from these structures . A cytotoxic attack toward EBV infected cells in the MS brain is consistent with enrichment in EBV-specific CD8+ T cells in the cerebrospinal fluid ( CSF ) of patients with early MS [53] , with increased CSF levels of granzymes during relapse in relapsing-remitting MS patients [54] , and with preferential expansion of CD8+ T cells in MS brain lesions and CSF [55] , [56] . Defects in the control of viral infections are suspected to promote the development of autoimmune diseases [57] . Nearly all of the genes whose variants have been associated with the risk of developing MS are implicated in immune system function [2] , making it plausible that in susceptible individuals subtle differences in the regulation of the immune response might allow an EBV infection to be established in the CNS and become the target of an immunopathological response . Experimental studies suggest that upon infection with persistent viruses establishment of extralymphatic viral sanctuaries depends both on organ anatomy and defective synergies between CD8+ T-cell- and antibody-mediated immune responses [58] . The present results do not answer the question of whether EBV dysregulation is consequence or cause of MS but disclose a link between EBV reactivation , antiviral immune response and disease activity during the relapsing-remitting stage of MS . Such a scenario is consistent with the results of randomized , double-blind , placebo-controlled clinical and MRI studies of anti-herpesvirus therapy in relapsing-remitting MS showing that anti-herpesvirus drugs inhibiting viral replication have beneficial effects in subgroups of patients with higher exacerbation rates and more severe disease activity [59] , [60] . Further work is needed to better understand whether and how an altered balance between EBV and the host immune system contributes to MS onset and verify the potential benefits of new antiviral drugs in controlling MS [61] .
All blood samples were obtained following acquisition of the study participants' written informed consent . The study protocol was reviewed and approved by the local ethics committes of S . Camillo Forlanini Hospital , Tor Vergata University , S . Andrea Hospital , and Fondazione S . Lucia . Use of post-mortem human brain material for the study purposes has been approved by the ethics committee of the Istituto Superiore di Sanità . MS patients and HD were recruited between 2008 and 2012 from Tor Vergata University , S . Camillo Forlanini and S . Andrea Hospitals in Rome . We enrolled 250 patients who were diagnosed the relapsing-remitting form of MS according to the 2005 revised McDonald's criteria [62] . A neurologist ( SR , CG , FB , DC , MS ) examined the patients , including assessment of the EDSS and confirmation of clinical relapse or remission . Of the 250 enrolled subjects , 113 MS patients were selected for this study according to their HLA genotype ( HLA-A*02101 , B*0801 ) for which well characterized EBV and CMV peptide antigens have been described [38]–[40] . Seventy-nine patients were free of therapy for at least 3 months , 20 patients were treated with IFN-β subcutaneously ( 12 with IFN-β 1a and 8 with IFN-β 1b ) for 1–11 years ( median = 4 years ) and 14 MS patients were treated with natalizumab for 8–16 months ( median = 12 months ) . The control subject group included 43 HD matched for sex and age and selected for their HLA genotype ( HLA-A*02101 , B*0801 ) . The demographic and clinical characteristics of HD and MS patients are summarized in Table 1 . At the time of peripheral blood collection 38 and 75 MS patients were in the active and inactive phase of the disease , respectively , based on clinical assessment and brain MRI ( Table 2 ) . Four MS patients ( 2 untreated and 2 treated with natalizumab ) and 3 HD were monitored longitudinally for 7–27 months and blood was drawn every 3 to 6 months . Seventy-six % ( 60/79 ) of untreated patients and all IFN-β-treated patients were examined by brain MRI with gadolinium enhancement on the same day or within 1 week before blood collection; only in one case MRI was performed 3 weeks before blood collection . All natalizumab-treated patients were monitored with MRI every 6 months . Acquisition of brain MRI scans was obtained in a single session . Conventional MRI scans were acquired including the following sequences: Fast Fluid Attenuated Inversion Recovery ( FLAIR ) , T1 weighted images ( T1-WI ) before and after gadolinium administration covering the whole brain . The gadolinium enhanced T1-WI scans were obtained for all patients 15 minutes after admnistration of gadolinium ( 0 , 1 mmol/kg ) . MRI scans were classified as active if there was at least one gadolinium enhancing lesion . As shown in Table 2 , the majority ( 86% ) of active MS patients included in this study had both clinical manifestations and an active MRI scan , while a minority showed either clinical ( 6% ) or MRI ( 8% ) evidence of disease activity . Conversely , all inactive patients exhibited neither clinical manifestations nor disease activity on MRI . PBMCs were isolated on a Ficoll gradient ( Ficoll-Paque PLUS , GE Healthcare ) and stained with pre-titrated antibodies . To evaluate the CD8+ T cell response to EBV latent and lytic antigens , PBMC from MS patients and HD were stained with fluorochrome-labeled pentamers ( ProImmune , Oxford , UK ) . The analysis was conducted on freshly isolated PBMC with the exclusion of dead cells , providing an accurate pattern of CD8+ T cell in vivo specificities . We analyzed CD8+ T cells specific for two EBV lytic protein epitopes , the HLA-A*0201 restricted epitope ( GLCTLVAML ) from BMLF-1 and the HLA-B*0801 restricted epitope ( RAKFKQLL ) from BZLF-1 , and for two EBV latent protein epitopes , the HLA-A*0201 restricted epitope ( CLGGLLTMV ) from LMP-2A and the HLA-B*0801 restricted epitope ( FLRGRAYGL ) from EBNA-3A . The CD8+ T cell response to an HLA-A*0201 restricted immunodominant peptide ( NLVPMVATV ) from pp65 of human cytomegalovirus ( CMV ) was studied as a control for anti-viral MHC class I restricted CD8 T-cell responses . One x 106 PBMCs were stained with 10 µl of PE conjugated-pentamers alone , washed with PBS and then stained with anti human CD3 APC Alexa e780 ( eBioscience Inc . , San Diego , CA ) and CD8 ECD ( Beckman Coulter , Brea , CA ) . Cells were also stained for dead cell exclusion ( Fixable Dead Cell Stain Kits , Invitrogen , Life Technologies , Paisley , UK ) . The samples were acquired on a CyAN ADP cytometer ( Beckman Coulter ) and analysed by FlowJo software ( Tree Star , Ashland , OR ) . Frequencies of pentamer+ cells below 0 . 02% of CD3+ T cells were considered as background staining as indicated by the manufacturer . An example of the gating strategy used to identify pentamer+ cells is shown in Figure S1 . Thirteen cerebral tissue blocks from 7 MS cases ( MS79 , MS92 , MS121 , MS154 , MS180 , MS234 , MS342 ) who died in the secondary progressive phase of MS and were characterized by substantial brain inflammation were analyzed in this study . Tissues were provided by the UK Multiple Sclerosis Tissue Bank at Imperial College in London after collection via a prospective donor program with fully informed consent . Based on the available clinical documentation , all MS patients were in the progressive phase of the disease , and no immunotherapy is reported in the 6 months before death . Control tissues for BZLF-1 immunohistochemistry included fixed-frozen brain sections from one control subject who died for cardiac failure ( obtained from the UK MS Tissue Bank ) , paraffin sections of a brain with tuberculous meningo-encephalitis and of an EBV-negative brain B-cell lymphoma ( kindly provided by Dr R . Hoftberger , Clinical Institute of Neurology , Wien ) , and paraffin sections of a tonsil from a subject with active infectious mononucleosis ( kindly provided by Dr G . Niedobitek , Sana Klinikum Lichtenberg/Unfallkrankenhaus , Berlin ) . Eight brain tissue blocks ( 4 cm3 each; 1 snap frozen , 7 fixed frozen ) from 5 MS cases ( MS92 , MS121 , MS154 , MS180 , MS342 ) were used for immunohistochemical studies . Lesion inflammatory activity was assessed as previously described [63] . Five snap-frozen brain tissue blocks from 4 MS cases ( MS79 , MS92 , MS180 , MS342 ) were used to study BZLF-1 gene expression using quantitative real-time RT-PCR . One snap-frozen control lymph node was obtained from Dr Egidio Stigliano , Policlinico A . Gemelli , Rome . Brain sections were stained using immunohistochemical and single or double indirect immunofluorescence techniques . Immunohistochemical detection of CD20 , MHC class II antigen and myelin-oligodendrocyte glycoprotein ( MOG ) , and immunofluorescence stainings for BFRF1 , Ig-A , -G , -M , CD8 and perforin alone or in different combinations were performed as previously described [32] . For BZLF-1 immunohistochemistry , deparaffinised sections from infectious mononucleosis tonsil , cerebral B cell lymphoma and brain with tuberculous meningo-encephalitis and cryosections from PFA-fixed control brain were subjected to antigen retrieval procedure in citrate buffer in microwave for 3 cycles of 3 min each before quenching of endogenous peroxidase activity in PBS containing 0 , 1% H2O2 . Sections were treated with 0 . 5% Triton X-100 in PBS for 10 minutes and incubated for 1 hour with normal serum 30%+0 , 25% Triton X-100 and then overnight at 4°C with mouse monoclonal antibody ( mAb ) specific for BZLF-1 protein ( clone BZ-1 , kindly provided by Dr J . Middeldorp , VUMC , Amsterdam ) diluted 1∶50 in PBS containing 0 , 25% Triton X-100 . Sections were then incubated with biotin-conjugated rabbit anti-mouse antibody ( Jackson Immunoreaearch Laboratories , West Grove , PA ) for 1 hour at room temperature ( RT ) , ABC-peroxidase complex ( Vector Laboratories , Burlingame , CA ) for 45 min , and AEC ( DakoCytomation , Glostrup , Denmark ) or diaminobenzidine ( Sigma , St Louis , MO ) to reveal the immune reaction . The EBV-producing B95-8 cells ( marmoset B-cell line transformed with EBV ) [64] were used as positive control for BZLF-1 immunofluorescence staining . Paraformaldehyde-fixed frozen brain sections from MS cases and one control case were air-dried and post-fixed in 4% PFA for 5 minutes at RT or in iced acetone for 10 minutes at 4°C . Sections were subjected to antigen unmasking , permeabilization steps and block of unspecific binding sites as described above and then incubated for 36 h at 4°C with BZ-1 mAb ( diluted 1∶50 in PBS +0 , 1% Triton X-100 ) . Antibody binding was visualized using tetramethyl rhodamine isothiocyanate ( TRITC ) -conjugated goat anti-mouse antibody ( Jackson Laboratories ) containing 5% normal goat serum for 50 minutes at RT . After washing , sections were sealed in DAPI-containing medium or incubated further with FITC-conjugated rabbit anti-human Ig-A-G-M ( 1∶400; Dako Cytomation ) for 1 hour at RT . For double immunofluorescence for BZLF-1 and CD8 , sections were stained with BZ-1 mAb and anti-human CD8 rabbit polyclonal antibody ( 1∶50; Pierce , Thermo Fisher Scientific Inc . Rockford , IL ) followed by a mixture of Alexa Fluor 488-conjugated goat anti-mouse and TRITC-conjugated goat anti-rabbit secondary antibodies . Double and triple immunostainings for CD8/granzyme B and CD8/granzyme B/BFRF1 were performed by incubating PFA-fixed cryosections with a mixture of anti-CD8 rabbit polyclonal antibody and anti-granzyme B mAb ( Dako ) , or anti-CD8 mAb , anti-BFRF1 rabbit polyclonal ( 1∶800 ) and anti-granzyme B goat polyclonal ( 1∶50 , Santa Cruz ) antibodies overnight at 4°C , and then with a mixture of donkey FITC anti-mouse ( Invitrogen , Eugene , OR ) , TRITC anti-rabbit and AMCA anti-goat ( Jackson Immunoresearch Lab ) secondary antibodies . Sections were sealed in ProLong Gold antifade reagent with 4′ , 6′-diamidino-2-phenylindole ( DAPI ) ( Invitrogen ) or in Vectashield ( Vector Laboratories ) . Images were analysed and acquired with a digital epifluorescence microscope ( Leica Microsystem , Wetzlar , Germany ) . Negative control stainings were performed using Ig isotype controls and/or pre-immune sera . Snap-frozen brain tissue blocks from 4 MS cases ( MS79 , MS92 , MS180 , MS342 ) and control lymph node were used for laser capture microdissection and subsequent RNA analysis . For each tissue block , the integrity and quality of total RNA extracted with the SV Total RNA Isolation System ( Promega , Madison , WI ) were checked on ethidium bromide containing 1% agarose gels in Tris-borate/EDTA buffer . Ten to 20 serial brain sections for each MS case and from control lymph-node were mounted on membrane slides for laser capture microdissection ( MMI AG , Glattbrugg , Switzerland ) and processed as described previously [33] . Sections before and after these series were stained for CD20 and Ig to identify B cell- and plasma cell-containing regions in the inflamed meninges and white matter lesions . Using a laser microdissector SL Cut ( MMI AG ) equipped with a Nikon Eclipse TE2000-S microscope , we isolated areas containing meningeal infiltrates and B-cell follicles , lesioned grey matter , B cell-enriched perivascular cuffs in white matter lesions , lesioned white matter surrounding inflamed blood vessels , and normal-appearing white matter . The same brain areas were cut in 3 to 10 serial sections and pooled in a single cap . B cell follicles were isolated from the lymph-node . The isolated tissue fragments were collected in 50 µL of lysis buffer ( PicoPure RNA isolation kit , Arcturus Engineering ) , incubated at 42C° for 30 minutes and centrifuged at 800× g for 2 minutes . Lysates were stored at −80 C° until use . DNase-treated total RNA was extracted from 20-µm-thick brain sections or microdissected areas from 4 MS brains and 1 control lymph node , as previously described [32] . RNA samples were reverse-transcribed with oligo ( dT ) and random hexamers using the Murine Leukemia Virus Reverse Transcriptase ( Invitrogen Life Technologies , Carlsbad , CA ) . PreAMP Master Mix Kit ( Applied Biosystems , Foster City , CA ) was used to enrich for both cellular and viral gene transcripts . The cDNAs obtained from whole brain sections and microdissected brain and lymph node samples were preamplified according to the manufacturers' instructions using 90 nmol/L of each primer in a mix containing the same forward and reverse primers for GAPDH , CD19 and BZLF-1 used for real time RT-PCR . Quantitative PCR assays were performed in triplicate , as previously described [33] . cDNA from EBV-positive P3HR-1 cells and human primary B cells were included in each run as positive controls for BZLF-1 and CD19 gene expression , respectively . Sample values were normalized by calculating the relative quantity of each mRNA to that of GAPDH using the formula 2−ΔCt , where ΔCt represents the difference in cycle threshold ( Ct ) between target mRNA and GAPDH mRNA . The following primer pairs were used in this study: GAPDH_for ACAGTCCATGCCATCACTGCC; GAPDH_rev GCCTGCTTCACCACCTTCTTG [33]; BZLF-1_for GTTGTGGTTTCCGTGTGC; BZLF-1_rev AACAGCTAGCAGACATTGGTG [65]; CD19_for AGAACCAGTACGGGAACGTG; CD19_rev CTGCTCGGGTTTCCATAAGA [33] . Differences between categorical variables were evaluated by Pearson's chi-squared test while differences between continuous variables were analysed by unpaired t-test with 95% confidence intervals . | There is general consensus that multiple sclerosis ( MS ) is associated with Epstein-Barr virus ( EBV ) infection but the mechanistic links are still debated . EBV is a B-lymphotropic herpesvirus widespread in the human population and normally contained as a persistent , asymptomatic infection by immune surveillance . However , EBV can cause infectious mononucleosis , is associated with numerous human malignancies , and is implicated in some common autoimmune diseases . While EBV infection alone cannot explain MS development , it has been postulated that in susceptible individuals alterations in the mechanisms regulating the immune response to the virus may contribute to MS pathogenesis . Here , we show that MS patients with inactive disease exhibit a lower CD8+ T-cell response to EBV when compared to healthy donors and active MS patients while the latter have a higher frequency of CD8+ T cells specific for EBV lytic antigens . Therapy with interferon-β and natalizumab , two treatments for relapsing-remitting MS , was associated with marked changes in the EBV specific CD8+ T cell response . We also demonstrate that one of the EBV lytic antigens recognized by CD8+ T cells expanding in the blood during active MS is expressed in the inflamed MS brain . Our results support a model of MS pathogenesis in which EBV infection and reactivation in the CNS stimulates an immunopathological response and suggest that antiviral or immunomodulatory therapies aimed at restoring the host-EBV balance could be beneficial to MS patients . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunopathology",
"virology",
"immunology",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"immune",
"response",
"pathogenesis"
] | 2013 | Increased CD8+ T Cell Response to Epstein-Barr Virus Lytic Antigens in the Active Phase of Multiple Sclerosis |
The exon junction complex ( EJC ) is an RNA binding complex comprised of the core components Magoh , Rbm8a , and Eif4a3 . Human mutations in EJC components cause neurodevelopmental pathologies . Further , mice heterozygous for either Magoh or Rbm8a exhibit aberrant neurogenesis and microcephaly . Yet despite the requirement of these genes for neurodevelopment , the pathogenic mechanisms linking EJC dysfunction to microcephaly remain poorly understood . Here we employ mouse genetics , transcriptomic and proteomic analyses to demonstrate that haploinsufficiency for each of the 3 core EJC components causes microcephaly via converging regulation of p53 signaling . Using a new conditional allele , we first show that Eif4a3 haploinsufficiency phenocopies aberrant neurogenesis and microcephaly of Magoh and Rbm8a mutant mice . Transcriptomic and proteomic analyses of embryonic brains at the onset of neurogenesis identifies common pathways altered in each of the 3 EJC mutants , including ribosome , proteasome , and p53 signaling components . We further demonstrate all 3 mutants exhibit defective splicing of RNA regulatory proteins , implying an EJC dependent RNA regulatory network that fine-tunes gene expression . Finally , we show that genetic ablation of one downstream pathway , p53 , significantly rescues microcephaly of all 3 EJC mutants . This implicates p53 activation as a major node of neurodevelopmental pathogenesis following EJC impairment . Altogether our study reveals new mechanisms to help explain how EJC mutations influence neurogenesis and underlie neurodevelopmental disease .
Proper function of the cerebral cortex , our brain structure responsible for higher cognitive functions , relies upon embryonic neurogenesis . During neurogenesis , neural stem cells ( NSCs ) generate excitatory neurons [1 , 2] . In mice the onset of neurogenesis is embryonic day ( E ) 10 . 5 , when NSCs consist of neuroepithelial cells that primarily undergo self-renewal divisions . As development proceeds , neuroepithelial cells are replaced by radial glial cells that generate neurons either directly , or indirectly via new NSCs and intermediate progenitors ( IPs ) ( Fig 1A ) [3–5] . Defective neurogenesis impacts neuron production and can cause neurodevelopmental disorders such as microcephaly , in which brain size is severely reduced . To elucidate causes for such diseases requires a comprehensive understanding of how NSCs mediate proper brain development . One level of control increasingly implicated in NSC function and disease is post-transcriptional regulation [6–8] . In particular , a set of RNA binding proteins associated with developmental pathologies of the cerebral cortex is the exon junction complex ( EJC ) . The core EJC , composed of Rbm8a ( Y14 ) , Magoh , and Eif4a3 ( Ddx48 ) , influences mRNA splicing , translation , mRNA localization , and nonsense mediated decay ( NMD ) , via direct interactions with both RNA and auxiliary proteins in the nucleus and cytoplasm [9–15] . Copy number variations of RBM8A , EIF4A3 , and peripheral EJC components , are each strongly associated with neurodevelopmental phenotypes [16–18] . Moreover RBM8A and EIF4A3 mutations cause TAR syndrome and Richieri-Costa-Pereira syndrome , respectively , both of which are associated with neurological deficits [19–22] . While altered EJC levels are significantly linked to neurodevelopmental diseases , the pathogenic mechanisms by which EJC impairment causes these disorders are largely unknown . Recent studies from our lab have helped shed light on this question , with the discoveries that haploinsufficiency for either Magoh or Rbm8a , disrupts mouse cortical development . In these mouse models , both NSCs and IPs are depleted , neurons are ectopic , and there is massive apoptosis of neurons and progenitors , all leading to severe microcephaly [23–26] . We recently discovered that in Magoh mutants these neurogenesis phenotypes may be due in part to prolonged mitosis of NSCs [26] . Moreover , we identified Lis1 as one relevant Magoh downstream target during neurogenesis [23] . While these studies show Magoh and Rbm8a are essential for corticogenesis , it remains unknown if impairment of the third major EJC constituent , Eif4a3 , causes microcephaly . Additionally , if all EJC components are required in the developing brain , it is unclear whether they function via common regulatory pathways . This information is critical to understand how EJC genes regulate brain development . In this study we examined mice haploinsufficient for Magoh , Rbm8a , or Eif4a3 , to expose mechanisms by which EJC dysfunction impacts cortical development . First , we generated a NSC-specific conditional Eif4a3 mouse model to demonstrate that Eif4a3 haploinsufficiency phenocopies the aberrant neurogenesis and microcephaly seen in Rbm8a and Magoh mutants . We then utilized transcriptomic and proteomic analyses to uncover common genetic pathways controlled by all 3 EJC components at the onset of neurogenesis . These include expression of factors associated with the ribosome , proteasome , and p53 signaling pathway . All 3 EJC mutants showed splicing alterations in RNA processing factors , implicating the EJC in regulating a network of RNA metabolism factors . Finally , we focus on one of these downstream pathways , p53 , and show that p53 ablation significantly rescues microcephaly of all 3 EJC mutants . Altogether our study reveals novel mechanisms to help explain how EJC deficiency disrupts neurogenesis , implicating elevated p53 signaling in the etiology of EJC-mediated neurodevelopmental pathologies .
We previously showed that NSC-specific haploinsufficiency for either Magoh or Rbm8a causes microcephaly in mice [23–25] . To understand whether common mechanisms contribute to microcephaly following depletion of EJC core components , we first sought to address the role of the third core EJC component , Eif4a3 , in brain development ( Fig 1B ) . We examined the expression profile of Eif4a3 relative to Magoh and Rbm8a at early stages of cortical development . RT-qPCR showed that Magoh , Eif4a3 , and Rbm8a are expressed in the developing neocortex and show parallel increases in expression as neurogenesis proceeds ( Fig 1C ) . In situ hybridization revealed enriched Eif4a3 expression in the proliferative ventricular and sub-ventricular zones of the E14 . 5 neocortex , where NSCs reside , in a similar pattern to Rbm8a and Magoh [23 , 24 , 27] ( S1A Fig ) . Immunostaining showed that at the onset of neurogenesis ( E10 . 5 ) , EIF4A3 protein is expressed at detectable levels and is primarily localized within the nucleus , similar to MAGOH and RBM8A ( Fig 1D–1I ) . Together , these analyses indicate that Eif4a3 , Magoh and Rbm8a are co-expressed spatially and temporally in the developing mouse neocortex . We generated a conditional mouse carrying a floxed allele of Eif4a3 ( Eif4a3lox/+ ) to assess the phenotype of Eif4a3 deficiency in the developing brain ( Fig 2A ) . Eif4a3lox/+ mice were crossed to Emx1-Cre , which drives Cre expression in NSCs of the dorsal neocortex beginning at E9 . 5 [28 , 29] ( cre . jax . org ) . Genotyping of genomic DNA from Emx1-Cre;Eif4a3lox/+ mice confirmed the presence of predicted bands for both wildtype and lox alleles ( S1B Fig ) . Following Cre recombination , exon 2 is excised to generate a transcript predicted to undergo NMD-mediated degradation . Consistent with this , Eif4a3 mRNA and protein were reduced by about 50% in Emx1-Cre;Eif4a3lox/+ neocortices ( Fig 2B and 2C ) . These data demonstrate Eif4a3 can be efficiently depleted in the conditional haploinsufficient mouse model . We next analyzed the impact of Eif4a3 haploinsufficiency upon neurogenesis . At E12 . 5 , Emx1-Cre;Eif4a3lox/+ cortices were markedly smaller at a whole mount level ( Fig 2D and 2E ) and 30% thinner when compared to control ( Emx1-Cre ) littermates ( Fig 2F ) . PAX6-positive NSCs were significantly reduced in density in Emx1-Cre;Eif4a3lox/+ neocortices compared to control , an observation corroborated by western analysis ( Fig 2G , 2I , 2K and S1C and S1D Fig ) . The depletion of NSCs was associated with a concomitant increased thickness of the TUJ1-positive neuronal layer ( Fig 2H and 2J ) . These findings , smaller brain size , NSC depletion , and excessive neurons , demonstrate that Eif4a3 haploinsufficiency phenocopies Emx1-Cre;Rbm8alox/+ and Emx1-Cre;Magohlox/+ neocortices [23 , 24] . We also previously showed that MagohMos2/+ germline mutant and Emx1-Cre;Rbm8alox/+ NSCs exhibit significant mitotic defects [23 , 24 , 26] . Quantification of mitotic index using phospho-histone 3 ( PH3 ) staining revealed increased mitotic index of E11 . 5 Emx1-Cre;Eif4a3lox/+ neocortices compared to control ( Fig 2L–2N ) . Extensive apoptosis as evidenced by cleaved-caspase3 ( CC3 ) , is also associated with Magoh and Rbm8a haploinsufficient brains [23 , 24] . Similarly , we noted extensive apoptosis in Emx1-Cre;Eif4a3lox/+ neocortices ( Fig 2O and 2P ) . Apoptosis was present in both neurons and NSCs throughout the dorsal cortex , similar to Magoh and Rbm8a mutants ( Fig 2Q–2T ) . These early neurogenic phenotypes impacted brain structure . At E14 . 5 , the dorsal telencephalon was largely absent in Eif4a3 haploinsufficient brains ( Fig 2U and 2V ) . Of the remaining dorsal telencephalon tissue found adjacent to the pallial-subpallial boundary , the cortex was extremely thinned and neurons were disorganized ( Fig 2W and 2X ) . These phenotypes are highly similar to Emx1-Cre;Rbm8alox/+ brains [24] . Surprisingly , despite the prevalent disruption of the developing neocortex , Eif4a3 , Rbm8a , and Magoh conditional mutant mice survive into adulthood [24 , 25] . We evaluated postnatal ( P ) brain sizes of each EJC mutant . Comparison of P12 whole mount brains demonstrated significant reductions in all 3 EJC mutants ( Fig 3A–3E ) . Both Eif4a3 and Rbm8a haploinsufficiency caused severe microcephaly , with an average 70% reduction in cortical area of whole mount brains [24] ( Fig 3E ) . The microcephaly phenotype of Rbm8a and Eif4a3 mutant mice was significantly worse than Magoh haploinsufficient mice , which exhibited a 40% reduction [24 , 25] . This phenotypic difference may be due to redundant expression of a second Magoh homolog , whereas the other EJC components do not have identifiable homologs [30] . Together with our previous studies , these analyses indicate that Eif4a3 , Magoh , and Rbm8a each control similar aspects of neurogenesis ( NSC proliferation , number and apoptosis ) , and ultimately brain size . Given the overlapping expression patterns , common neurogenesis phenotypes , and vast literature connecting Magoh , Rbm8a , and Eif4a3 , we hypothesized that these EJC components work together to influence cortical development . To test this , we aimed to identify molecular changes associated with early neurogenesis in each of the three EJC mutants . We performed transcriptome profiling of E10 . 5 neocortices from the following genotypes: Emx1-Cre , Emx1-Cre;Rbm8alox/+ , Emx1-Cre;Magohlox/+ , and Emx1-Cre;Eif4a3lox/+ ( n = 3 biological replicates each ) ( Fig 4A ) . We focused on E10 . 5 for several reasons . This stage marks the beginning of neurogenesis when the neocortex is composed primarily of self-renewing neuroepithelial NSCs [4] . Moreover , it is just prior to the onset of severe defects in EJC mutants , and a stage when all 3 genes are reduced in their respective mutants , as evidenced by RT-qPCR of the RNA-sequencing samples ( Fig 4B ) . We examined global RNA changes in the 3 mutants relative to the control and to each other . Amongst the 18 , 465 detectable coding and non-coding transcripts expressed in the E10 . 5 control cortex , 2 . 9% were altered in Emx1-Cre;Rbm8alox/+ , 0 . 9% were altered in Emx1-Cre;Eif4a3lox/+ , and 0 . 4% were altered in Emx1-Cre;Magohlox/+ ( FDR , q<0 . 05 ) ( S1 Table ) . Hierarchical clustering of these significantly altered transcripts revealed segregation of control and mutant biological replicates for all 3 EJC mutants , as evidenced in heat maps ( Fig 4C ) . Equivalent proportions of transcripts were upregulated and downregulated within individual EJC mutants ( Fig 4D , S1 Table ) . We validated expression for several differentially expressed transcripts , Tbr2 , Ngn2 , NeuroD6 , and Gtse1 , using RT-qPCR , which showed similar trends to RNA-seq data ( Fig 4E and 4F , S2B Fig ) . Despite the fact that the EJC binds a large fraction of expressed transcripts in immortalized cells [31–33] , these experiments suggest EJC haploinsufficiency does not broadly impair transcript levels of E10 . 5 neocortices . This observation echoes previous microarray studies of germline MagohMos2/+ mutant brains [23] , Eif4a3 silenced Xenopus [34] , and EJC Drosophila mutants [35] . We next assessed the extent to which transcripts overlapped amongst the EJC mutants , focusing only on the fraction of alterations which were highly significant ( q<0 . 05 ) . We noted extensive overlap in pairwise comparisons between individual mutants ( S2A Fig ) . Of the 70 Magoh dependent transcripts , 87% were altered in Rbm8a mutants and 46% were altered in Eif4a3 mutants . Of the 172 transcripts altered in Eif4a3 mutants , 19% overlapped with Magoh mutants and 46% overlapped with Rbm8a mutants . Fisher’s exact tests demonstrated these overlapping changes were highly significant . In all 3 mutants , 31 transcript changes overlapped , which represents 6% , 18% , and 44% of all altered transcripts in the Rbm8a , Eif4a3 , and Magoh mutants , respectively . As noted by Venn diagram , some transcript alterations were specific to individual mutants ( S2A Fig ) . This was especially evident in Rbm8a and Eif4a3 mutants , and suggests there could be roles for EJC components outside of the complex . Yet , taken together , these data support the notion that EJC components also work together to selectively affect mRNA levels at the onset of neurogenesis . Given the significant overlap in transcript changes , we hypothesized Magoh , Rbm8a , and Eif4a3 mutants influence common molecular and cellular pathways . To determine if this was true , we performed gene set enrichment assays ( GSEA ) Kyoto Encyclopedia and Genes and Genomes ( KEGG ) analysis on all 18 , 465 detectable genes from the transcriptome data , ranked by p value . Of those pathways significantly altered in all 3 EJC mutants , we discovered enrichment in only ribosome , proteasome , and p53 signaling ( Fig 4G and S2 Table ) . This was also evidenced by enrichment plots and STRING analyses ( S3A–S3I Fig ) . Closer inspection of the significantly altered transcripts within each KEGG category revealed extensive overlap amongst the 3 mutants ( S3B , S3D and S3F Fig ) . Gene ontology ( GO ) analyses using GSEA further corroborated ribosomal alterations in all 3 haploinsufficient mutants ( S2C Fig and S2 Table ) . The directionality and degree of expression changes in ribosome-encoding transcripts were consistent across all mutant mice ( S2D Fig ) . Notably , inspection of only the significant transcript changes for each mutant ( q<0 . 05 ) showed that ribosomal-associated transcripts made up 11 . 4% , 6 . 5% and 7 . 5% of Magoh , Rbm8a , and Eif4a3 mutants , respectively . This indicates altered protein homeostasis pathways , including the ribosome , are shared early consequences of EJC haploinsufficiency . To assess transcript regulation in an independent EJC model not reliant on Cre , we performed RNA-sequencing on E10 . 5 neocortices from control ( C57BL/6J ) and germline Magoh haploinsufficient mice ( MagohMos2/+ ) ( n = 4 biological replicates each ) ( Fig 5A ) . Hierarchical analysis revealed consistent expression changes in MagohMos2/+ compared to control littermates ( Fig 5B ) . Amongst the 23 , 577 genes detected , only 80 ( 0 . 3% ) transcripts were differentially expressed ( q<0 . 05 ) , and these were equivalently upregulated and downregulated ( Fig 5C , S1 Table ) . RT-qPCR validation confirmed alterations in two transcripts , Dclk1 and Tbr2 , with similar trends to RNA-seq ( Fig 5D ) . Changes were more dramatic than in Emx1-Cre;Magohlox/+ , consistent with a more severe impact of the Magoh germline deletion [23 , 25] . GSEA KEGG analysis of all detectable transcripts revealed significant enrichment for ribosome , proteasome , and p53 signaling components , amongst additional regulators of protein metabolism ( Fig 5E and S4A Fig ) . GO analysis also detected ribosomes as a top altered category ( S4C Fig ) . Of note , we observed overlap between Emx1-Cre;Magohlox/+ and MagohMos2/+ transcripts within the ribosome , proteasome , and p53 categories ( S4B Fig ) . Altogether , these transcriptome analyses from 4 independent mouse lines , including 2 models of Magoh , demonstrate that EJC haploinsufficiency influences a few common pathways including ribosome , proteosome and p53 signaling . Given the requirement of the EJC in splicing and NMD , we next assessed splicing isoforms in the transcriptome data . Consistent with a published study in human cell lines [36] , widespread splicing changes were evident in all 3 EJC mutants compared to control ( S3 Table ) . We measured specific splicing events relative to all annotated alternative splicing ( AS ) events using Mixture-of-Isoforms ( MISO ) software [37 , 38] . Comparing the changed AS events to all annotated AS events , the distribution of AS types was significantly altered , with a 2–3 fold enrichment in retained intron ( RI ) events in all 3 EJC mutants ( p<0 . 001 ) ( Fig 6A ) . Amongst the RI events , 61% , 70% , and 23% were increased in Magoh , Rbm8a , and Eif4a3 haploinsufficient mutants , respectively ( Fig 6B , S3 Table ) . In Emx1-Cre;Rbm8alox/+ , 91% of RI events introduced a premature stop codon , which presumably leads to mRNA degradation through NMD ( S3 Table ) . We validated several events , including Mapk13 in E11 . 5 Emx1-Cre;Rbm8alox/+ brains and Fus in MagohMos2/+ brains , noting alterations consistent with predictions ( Fig 6E and S5A Fig ) . Thus , the enrichment of RI events could be due to inefficient NMD activity [12 , 14] . Consistent with previous findings that EJC Drosophila mutants cause increased RI events [39–41] , our data suggest EJC components influence mRNA splicing in NSCs . We next used bioinformatics analysis to determine if there are overlapping classes of splicing variants in the 3 EJC mutants . We performed KEGG analysis on those genes with significant alterations in splice variant expression ( Bayes factor > 20 ) using the Database for Annotation , Visualization and Integrated Discovery ( DAVID ) . These analyses showed common terms amongst all EJC mutants , including a significant enrichment of spliceosome ( Fig 6F ) . GO analysis reinforced this finding , with enrichment of RNA regulatory categories including ribonucleoproteins and ribosomes ( S4 Table ) . 51 identical alternative splicing events were predicted among all 3 EJC mutants . String analyses of these genes revealed two clusters for ribosome regulation and splicing regulation ( S5B Fig ) . These data suggest that in addition to influencing transcript expression , EJC components have been co-opted to impact splicing of RNA regulators . Together , this implies an EJC dependent regulatory network that fine-tunes gene expression at the RNA level . We next measured the proteomes of control , Magoh , Rbm8a , and Eif4a3 haploinsufficient E11 . 5 neocortices using quantitative proteomic liquid chromatography/mass spectrometry ( LC_MS/MS ) analyses ( n = 3 biological replicates each ) ( Fig 7A , S5 Table ) . We detected 3 , 587 proteins in the control and assessed relative levels of these proteins in each of the mutants . Magoh , Eif4a3 , and Rbm8a haploinsufficiency led to significant alterations in 3 . 8% , 1 . 5% , and 4 . 3% of the detectable proteome , respectively ( p<0 . 05 ) . Consistent with our transcriptome analysis , the proteomes of the various mutants showed both overlapping and independent alterations ( S6A Fig ) . We next asked if there were common alterations amongst those proteins significantly altered in the 3 EJC mutants . Using KEGG DAVID analysis to examine only significant protein changes ( p<0 . 05 ) , we identified ribosomes as the only pathway enriched in all 3 EJC mutants , significant in 2 of the mutants ( Fig 7B ) . GO analysis showed components of ribosomes and ribonucleoprotein complexes amongst the most significantly enriched categories ( Fig 7C , S6 Table ) . We performed STRING analysis of all altered proteins in the EJC mutants within the largest GO term , “Ribonucleoprotein Complex , ” which included ribosome components and splicing factors ( Fig 7D ) . This analysis reinforced strong regulatory networks present amongst proteins downstream of the EJC , and the consistent directional changes evident in all 3 mutants . Closer inspection of all significant protein changes within each mutant showed that ribosomal proteins made up 7 . 9% , 5 . 8% , and 7 . 5% of Magoh , Rbm8a , and Eif4a3 mutant changes , respectively . A large fraction of ribosomal proteins changed consistently across all 3 mutants , showing up and down regulation at the protein level ( S6B Fig ) . Altogether these genomic and proteomic analyses support the notion that ribosome and ribonucleoprotein alterations are major early defects associated with EJC deficiency in the developing brain . The omics analyses pointed to several common pathways that are dysregulated at the onset of neurogenesis , and suggested candidate molecules that could be relevant for EJC mutant phenotypes . We hypothesized that p53 signaling , in particular , was a major contributor to EJC-mediated microcephaly . Activated p53 is a key regulator of apoptosis and defective cell cycle progression [42] , two major phenotypes of EJC mutant brains . Moreover , p53 target transcripts were upregulated in all 3 conditional EJC mutants and Magoh germline mutant ( Figs 4 and 5 , S1 Table ) . Additionally , a correlation has been previously observed in p53 transcript changes in Magoh germline brains and induced radiation [43] . Altogether these data suggest p53 activation may be a common critical node in disease pathogenesis following EJC impairment . We thus probed the relationship between EJC haploinsufficiency and p53 signaling , by assessing p53 nuclear accumulation in embryonic brain sections , as a proxy for pathway activation [26] . Haploinsufficiency for Magoh , Rbm8a , and Eif4a3 led to a significant accumulation of p53 in the VZ compared to control brains , which showed no evidence of p53 accumulation ( Fig 8A–8J ) . Western blotting confirmed accumulation of p53 protein in Eif4a3 mutant cortices ( S7A Fig ) . P53 activation was evident in E11 . 5 Rbm8a mutants ( Fig 8E and 8F ) , prior to the onset of apoptosis [24] , and was specifically enriched in PAX6-positive NSCs ( Fig 8I and 8J ) . This demonstrates that p53 is activated in EJC haploinsufficient NSCs . We hypothesized that p53 activation contributes to microcephaly phenotypes of all 3 EJC mutants . To directly assess this , we crossed Emx1-Cre;Magohlox/+ , Emx1-Cre;Rbm8alox/+ , and Emx1-Cre;Eif4a3lox/+ , onto a p53lox/lox null background . We collected E18 . 5 embryos and measured cortical area . Compared to control , Emx1-Cre;p53lox/lox did not alter brain size ( Fig 9A , 9C , 9F , 9H , 9K and 9M ) . As expected , cortical area was significantly reduced in mice haploinsufficient for Magoh , Rbm8a , or Eif4a3 , to a similar degree seen in adults ( Compare Fig 9B , 9G and 9L to Fig 3 ) . Strikingly , for all 3 EJC mutants the microcephaly was significantly , albeit partially , rescued in a p53 mutant background ( Fig 9D , 9I and 9N ) . Amongst the 3 EJC mutants , the extent of p53-mediated rescue varied and was most effective in the least severe microcephaly mutant , Magoh ( Fig 9E , 9J and 9O ) . These data indicate that p53 activation is a major cause of microcephaly in all 3 EJC mutants . Our data also suggest that for Rbm8a and Eif4a3 , additional p53 independent factors likely contribute to the reduced brain size . To elucidate the nature of the p53-mediated rescue we examined apoptosis and neuron number . Amongst the 3 core EJC components , reduced RBM8A levels are the most strongly associated with human microcephaly [19 , 21 , 44] . Given this clinical relevance , we focused our analysis on the Rbm8a mutant . As p53 is essential for induction of apoptosis , we first assayed whether p53 ablation rescued apoptosis in the Rbm8a mutant . As predicted , CC3 immunostaining revealed complete rescue of apoptosis in E12 . 5 Emx1-Cre;Rbm8alox/+;p53lox/lox brains ( S7B–S7D Fig ) . Thus p53 activation promotes apoptosis downstream of Rbm8a . We next examined neuronal layers of E18 . 5 brains ( Fig 10A–10D ) . As we have previously shown [24] , Rbm8a mutant brains are missing most of their pallium ( Fig 10B ) . In the Emx1-Cre;Rbm8alox/+;p53lox/lox brains , the pallium is restored , consistent with the rescue of apoptosis ( Fig 10D ) . We asked if p53 loss impacts neuronal layers , focusing on the tissue adjacent to the pallial-subpallial boundary which is still present in Rbm8a mutants [24] . We quantified both deep and superficial neuronal markers which are generated at early and late stages of neurogenesis , respectively [4] . As predicted , Cux1+ neurons ( layer II/III ) were nearly ablated in Emx1-Cre;Rbm8alox/+ brains , compared to control or p53 alone ( Fig 10E–10G ) . In contrast , in p53;Rbm8a compound mutant brains Cux1+ neuron number was largely rescued ( Fig 10H and 10M ) . Another marker of both superficial and some deep layer neurons , Satb2 , was reduced in Emx1-Cre;Rbm8alox/+ , but partially rescued in a p53 null background ( S7E–S7I Fig ) [45 , 46] . We also examined earlier born deep layer Tbr1+ neurons ( Fig 10I–10L ) . As we previously described [24] , in Emx1-Cre;Rbm8alox/+ brains Tbr1 number is normal but distribution is skewed basally ( Fig 10I , 10J , 10K and 10N–10P ) . This is consistent with our previous finding that at early stages of development , Tbr1 density is increased in Rbm8a mutants , perhaps due to increased neuron production [24] . In Emx1-Cre;Rbm8alox/+;p53lox/lox brains , aberrant Tbr1+ neuron distribution was restored to normal ( Fig 10L , 10O and 10P ) . These analyses show that in Rbm8a mutants , p53 activation influences the number and distribution of neurons generated at different stages of neurogenesis , and plays a particularly important role in genesis of upper layer neurons . Taken together , our data implicate p53 activation as a key node in the microcephaly pathology following EJC impairment .
Our study elucidates several layers of EJC-dependent gene expression in the developing neocortex . Whereas EJC-dependent targets are known in Drosophila and immortalized cells [35 , 36] , our study is the first to discover EJC-dependent gene expression in a mammalian stem cell population . We demonstrate that EJC haploinsufficiency alters only a small fraction of the transcriptome , and these changes are disproportionately enriched for ribosomal proteins , proteasome components , and p53 signaling . Thus , the EJC may be especially important in regulation of protein homeostasis machinery . We also find the 3 core EJC proteins converge in regulating alternative splicing events . In particular we identify aberrant intron retention events which are suggestive of roles in mRNA splicing and NMD , and are consistent with genomic studies of EJC depletion in Drosophila and mammalian cells [36 , 39–41] . Notably these splicing changes are enriched for both spliceosomal and ribosomal components . Alterations in ribosomes are also observed at the protein level . Altogether , these analyses indicate the EJC is integral to an RNA regulation network controlling neurogenesis . These findings raise several fascinating questions . Although we focused on common EJC regulatory pathways , our data also highlight there are unique targets of individual EJC components . In future studies it will be of interest to consider potential independent roles for EJC components outside of the complex in neurogenesis . Another interesting question is how the EJC differentially regulates its targets in individual cells . For example , although we measured genomic changes in tissue that is mainly composed of 1 cell type , neuroepithelial progenitors , observed transcript and splicing differences could be attributed to progenitors in different cell cycle states . Moreover , it will be of interest to determine if the same pathways are regulated by the EJC in non-Emx1-derived cell types . We demonstrate that EJC mutant mice all exhibit profound microcephaly , which is significantly rescued by p53 deletion . Detailed analysis of Rbm8a mutants reveal that p53 attenuation partially restores superficial neuron number and distribution of deep layer neurons . Thus , the dramatic loss of upper layers in Rbm8a mutants is due , in part , to p53 activation . At least 2 scenarios could explain this rescue . P53 induction of apoptosis may severely reduce both neuron and progenitor number , particularly at later stages when upper layers are produced . Aberrant p53 activation may also influence stem cell divisions and thus their progeny . Our lab previously showed increased mitotic index in Magoh and Rbm8a mutant NSCs [23 , 24 , 26] . Mitotically delayed Magoh mutant NSCs preferentially produce neurons and apoptotic progeny , at the expense of NSCs [26] . We find that p53 activation is evident at E10 . 5 , which precedes the onset of mitotic defects at E11 . 5 and E12 . 5 . Given this sequence of events , it is tempting to speculate that aberrant p53 activation may influence progenitor ( and ultimately neuron ) number by delaying mitosis . Future experiments will be useful for evaluating if this relationship is correlative or causal . How might p53 be activated by EJC dysfunction ? It is plausible that ribosomal alterations contribute to p53 activation , as evidenced in many examples from the literature for genes controlling ribosome biogenesis [42 , 47–51] . Alternatively , p53 could be activated independent of the ribosome , as seen in the pancreas [52] . The EJC could also directly regulate RNA metabolism of p53 pathway components , as has been observed in splicing of apoptotic regulators [53] . The mechanisms contributing to p53 activation in EJC models are a topic of future interest . For Eif4a3 and Rbm8a mutants , p53 rescue was incomplete , suggesting there must be additional EJC-dependent pathways mediating early stages of neurogenesis . Our analyses implicate several promising candidates . Reduced expression of canonical neurogenesis regulators , including Ngn2 , Tbr2 , and NeuroD6 , could contribute to cell fate changes in the neocortex . All 4 EJC mutants also showed alterations in components of the proteasome , indicating that the EJC could influence neurogenesis by regulating protein homeostasis . Our data also identify ribosomal alterations at the transcriptome , splicing , and proteomic level , suggesting ribosome regulation could contribute to EJC-dependent microcephaly . Indeed , human genetic studies suggest that ribosome biogenesis defects cause neurodevelopmental diseases [54] . Of the significant ( FDR<0 . 05 ) changes in 3 different mutants , ribosomal transcripts made up 5–11% , well above the fraction expressed in progenitors , a finding which is reinforced with unbiased GSEA analysis . How might the EJC influence expression of ribosomal components ? The EJC could directly regulate ribosome biogenesis . Indeed , ribosome biogenesis defects are seen in Fal1p S . cerevisiae mutants [55 , 56] and Eif4a3 siRNA-depleted mammalian cells [56–58] . Alternatively , highly expressed genes , which include both ribosomal and proteasome components , could be especially sensitive to EJC levels . Another possible explanation for our results is that ribosome alterations are an indirect result of overall cellular stress . This idea is supported by the observation that some ribosome-encoding transcripts are not altered in all 3 mutants . It is also notable that ribosomal transcripts at E10 . 5 were nearly universally upregulated , whereas one day later the proteins were differentially altered . This could be due to differences in RNA versus protein regulation or could suggest compensatory responses to restore ribosomal levels in the brain . Understanding the nature of how the EJC influences ribosome stoichiometry , and how this may influence microcephaly , will be an important question for the future . Mutations and copy number variations in core and peripheral EJC components are strongly associated with neurodevelopmental deficits in humans , yet the etiology of these pathologies is poorly understood . Microdeletions and duplications of a 15-gene locus containing RBM8A are associated with microcephaly , macrocephaly , autism , and epilepsy [19 , 20] . Compound inheritance of this deletion and a regulatory RBM8A mutation is responsible for TAR syndrome , a congenital malformation of blood and skeletal systems which can also present with neurological deficits [21] . Moreover , regulatory EIF4A3 mutations cause a craniofacial disorder presenting with learning and language disabilities [22] . Intriguingly , both craniofacial and neurodevelopmental anomalies are associated with disruption of p53 signaling and ribosomal impairments [49 , 50 , 59 , 60] . It is notable the EJC downstream splicing changes include several genes , such as RPL10 , which are mutated in patients with neurodevelopmental disorders [60] . Thus , it is interesting to consider whether some of the expression changes we have identified in mouse models may contribute to EJC disease etiology . Altogether , based on our discoveries , we propose aberrant p53 signaling contributes to the pathology of EJC related disorders and that modifications of p53 signaling may be of potential therapeutic interest . It is tempting to speculate that EJC diseases could be considered as ribosomopathies . Going forward , the EJC haploinsufficient mouse mutants we have generated provide valuable models for understanding the etiology of microcephaly and dissecting cell autonomous requirements in NSCs . Future studies using ubiquitous knockout of EJC components may help to further model other disease manifestations . In summary , our findings demonstrate new mechanisms to explain how EJC haploinsufficiency causes microcephaly , which has implications for understanding physiological functions of the EJC in the developing brain and in disease pathogenesis .
All experiments were performed in agreement with the guidelines from the Division of Laboratory Animal Resources from Duke University School of Medicine and IACUC . Plug dates were defined as embryonic day ( E ) 0 . 5 on the morning the plug was identified . The conditional targeting vector for ES cell targeting was designed and generated by the Transgenic Facility at Duke University Cancer center . Positive ES clones were selected by performing long-range PCR of both arms . For long-range PCR of 5’ arms , the following conditions were used: 94°C X 1 min ( 1X ) ; 98°C X 10 s , 60°C X 15 s , 68°C X 8 . 5 min ( 40X ) ; 72°C X 10 min . 5’ F1:GTCCCAGAAATATCAGTGAGAATC; 5’ R1:CTTGTCATCGTCGTCCTTGTAGTC . For long-range PCR of 3’ arms , the following conditions were used: 94°C X 2 min ( 1X ) ; 98°C X 10 s , 60°C X 15 s , 68°C X 2 . 5 min ( 40X ) ; 72°C X 10 min . Positive clones were electroporated into CD1 blastocysts , and the resulting chimeras were mated to C57BL/6J females to obtain germ-line transmission . For genotyping Eif4a3lox mice , the following conditions were used: 94°C X 3 min ( 1X ) ; 94°C X 15 s , 62°C X 20 s , 72°C X 30s ( 30X ) ; 72°C X 10 min ( 1X ) . 5’ forward: CTTGCAGTTGTCTTTCTGCGG; 3’ Reverse: CACACATGGCGATCCGCTCG . The following strain was acquired from Jackson labs: Emx1-Cre ( B6 . 129S2-Emx1tm1 ( cre ) Krj/J ) . E10 . 5 neocortices and E11 . 5 dorsal cortices were collected from Emx1-Cre , Emx1-Cre;Eif4a3lox/+ , Emx1-Cre;Rbm8alox/+ , and Emx1-Cre;Magohlox/+ mice and lysed in RIPA lysis buffer with protease inhibitors ( Pierce , Rockford , IL ) . Cortical lysates were run on 4–20% pre-casted SDS–Polyacrylamide gels ( Bio-Rad ) . For Pax6 and P53 blots , stain free gels were used for total protein normalization . Gels were transferred onto nitrocellulose membranes and blotted using the following primary antibodies: rabbit anti-Eif4a3 ( 1:200 , Santa Cruz ) , rabbit anti-Pax6 ( 1:1 , 000 , Millipore ) , rabbit anti-p53 ( 1:1 , 000 , Leica ) and mouse anti-α-Tubulin ( 1:10 , 000 , Sigma ) . Blots were developed using ECL reagent ( Pierce ) . Densitometry was performed using ImageJ . Final values were quantified by normalizing EJC protein levels to loading controls ( 1:10 , 000 , Tubulin , Sigma ) or UV-induced Stain-free pre-casted gel ( Bio-Rad ) , and analyzed for significance using a Student’s t test . For qPCRs , whole neocortices from E10 . 5 and dorsal neocortices of E11 . 5 , and E12 . 5 and E14 . 5 embryos were collected from C57BL/6J ( wild-type ) , Emx1-Cre , Emx1-Cre;Eif4a3lox/+ , Emx1-Cre;Rbm8alox/+ , and Emx1-Cre;Magohlox/+ embryos and RNA was extracted using Trizol reagent ( Invitrogen ) followed by the RNeasy kit ( Qiagen ) . cDNA was prepared according to the iScript kit ( Bio-Rad ) . qPCR was performed in triplicates using Taqman probes ( Life Technologies ) : Rbm8a ( Mm04214345_s1 ) , Eif4a3 ( Mm00836350_g1 ) , Magoh ( Mm00487546_m1 ) , Ngn2 ( Mm00437603_g1 ) , Tbr2 ( Mm01351984_m1 ) , Dclk1 ( Mm00444950_m1 ) and Gapdh ( 4352339E ) . Sybr Green iTaq ( Biorad ) was performed with primers designed for Gste1 ( 5’Forward-CCAGAGCAAAGAGGACCAAG and 3’ Reverse-CCGTGAGAACTTTGGGGTTA ) , NeuroD6 5’ Forward-GCCTCAATGATGCTCTGGACAA and 3’ Reverse- CTCTTGCCAATCCTCAGAATTTCAG ) , and β-Actin ( 5’ Forward- CCTTCTTGGGTATGGAATCCTG and 3’ Reverse- GTTGGCATAGAGGTCTTTACGG ) . For wild-type samples at different developmental stages , semi-quantitative qRT-PCR was performed . A standard curve was generated with a 5 serial 10-fold dilution of cDNA from an independent E14 . 5 wildtype embryo . Final values were normalized to Gapdh loading control . For E10 . 5 control and conditional mutant samples , comparative qRT-PCR was performed . Values were normalized to Gapdh control . For each genotype , 3 embryos were examined , a student’s t test was run to determine the significance . For all experiments , 3 biological samples for each genotype were used . Brains were fixed overnight in 4% paraformaldehyde ( PFA ) at 4°C , followed by submersion in 30% sucrose until sinking , as previously described [23] . Brain cryostat sections ( 20 μm ) were prepared and stored at -80°C until use . Sections were permeabilized with 0 . 25% TritonX-100 for 10 min and blocked with MOM block reagent ( Vector laboratories ) for 1 hour at room temperature ( RT ) . Sections were incubated with primary antibodies for 2 hours at RT or overnight at 4°C . Sections were then incubated in species appropriate secondary antibodies and Hoechst for 15 min at room temperature . The following primary antibodies were used: rabbit anti-Magoh ( 1:200 , Proteintech ) , rabbit anti-Rbm8a ( 1:200 , Bethyl ) , rabbit anti-Eif4a3 ( 1:200 , Bethyl ) , rabbit anti-CC3 ( 1:200; Cell Signaling ) , rabbit anti-Pax6 ( 1:1 , 000; Millipore ) ; rabbit anti-PH3 ( 1:200 , Millipore ) , rabbit anti-Satb2 ( 1:1000; Abcam ) , anti-Cux1 ( 1:250 , Santa Cruz ) ; mouse anti-Pax6 ( 1:50; DSHB ) ; rabbit anti-p53 ( 1:250 , Leica ) , mouse anti-TuJ1 ( 1:400; Covance ) . The following secondary antibodies were used: Alex Fluor 488 , Alex Fluor 568 , Alex Fluor 594 ( 1:200–400; Invitrogen ) and Hoechst ( 1:1000; Invitrogen ) . High magnification images were captured using a Zeiss Axio Observer Z . 1 microscope coupled with an apotome . Cortical thickness was measured with Zen software . Quantifications were performed using ImageJ . A minimum of 3 sections from anatomically comparable regions per embryo and 3 biological replicates from control and mutants were measured/quantified . Control and EJC mutant embryonic neocortices were dissected at E10 . 5 . Samples were flash-frozen in liquid nitrogen and stored at -80°C until further treatment . RNA was extracted with Trizol ( Invitrogen ) followed by micro-RNeasy kit ( Qiagen ) according to manufacturer’s protocol . The library was generated with Kapa stranded mRNA-seq Kit . The fragmented poly-A RNAs were sequenced using Illumina Hi-Seq 2000 double end sequencing with 100nt length . RNA-seq data was processed using the TrimGalore toolkit ( http://www . bioinformatics . babraham . ac . uk/projects/trim_galore ) which employs Cutadapt to trim low quality bases and Illumina sequencing adapters from the 3’ end of the reads [61] . Only pairs where both reads were 20 nt or longer were kept for further analysis . Reads were mapped to the NCBIM38r73 version of the mouse genome and transcriptome using the STAR RNA-seq alignment tool [62] . Reads were kept for subsequent analysis if they mapped to a single genomic location . Gene counts were compiled using the HTSeq tool ( http://www-huber . embl . de/users/anders/HTSeq/ ) . Only genes that had at least 10 reads in any given library were used in subsequent analysis . Normalization and differential expression was carried out using the EdgeR Bioconductor package with the R statistical programming environment [63] . The exact test method was used to identify differentially expressed genes between the different mouse genotypes . Inspection of reads using integrative genomics viewer ( IGV ) software confirmed altered regulation of pseudogenes . Heatmaps were prepared for z-score transformed normalized expression for genes with an FDR , q<5% . To calculate significant overlap for Venn diagrams the following criteria were used: genes must with a q<0 . 05 and using a Fisher’s Exact Test for overlap between any two conditions . For alternative splicing analysis , Mixture-of-isoforms ( MISO ) [38] model was used to analyze RNA-Seq data and estimate the percent of splicing isoforms ( Ψ values , for ‘Percent Spliced Isoform’ ) , and the differentially spliced events are identified using a stringent filter ( bayes-factor >20 ) . The program was run with pooled samples of 3 biological replicates to reduce sampling biases . Validation of RI events was performed by RT-PCR with cDNA prepared from E11 . 5 dorsal cortices of control and EJC mutant embryos . The following primers were used: Fus Ex6 Forward: GGCCAAGATCAGTCCTCTATGAGT , Fus Ex8 Reverse: CATGACGAGATCCTTGATCCCGA , Mapk13 Ex6 Forward: GCAACCTGGCTGTGAATGAA , and Mapk13 Ex7 reverse: CTGGTTGTAATGCATCCAGCTG . For bioinformatic analysis of RNAseq Gene Set Enrichment Analysis ( GSEA ) was performed by creating a pre-ranked list of all detected transcripts , ranked by 1 minus the p value . The ranked list was imported into the GSEA software ( GSEA v2 . 2 . 2 , Broad Institute ) and analyzed using the pre-ranked gene list function . The data bases used were KEGGv5 . 1 and GOv5 . 1 . Common GSEA terms were cross compared among the 3 mutant strains and plotted according to their normalized enrichment score . The statistical test utilized by GSEA is the Kolmogorov-Smirnov statistical test . Venn diagrams include all genes that were identified as enrichment by the GSEA analysis within a given KEGG term . For splicing and proteomics analysis , we determined enriched pathways by assessing only significant hits ( splicing analysis: MISO , Bayes factor >20 and proteomics , p<0 . 05 ) . DAVID Annotation , Visualization and Integrated Discovery v6 . 7 was used to analyze significant changes by KEGG and gene ontology ( GO ) analysis ( including biological process , molecular function , and cellular component ) . Significance of enrichment in GO term analyses was calculated using the p value function given from a modified Fisher’s exact test by the DAVID database . For splicing analysis , STRING ( Search Tool for the Retrieval of Interacting Genes/Proteins ) analysis was carried out with transcripts show significant splicing changes ( Bayes>20 ) in all 3 EJC mutants . For proteomic analysis , STRING was carried out with significantly changed ( p<0 . 05 ) proteins in the “ribonucleoprotein complex” GO term . All components not connected to other genes/proteins were not included in figures . We performed relative quantitation proteomic study using the Duke Proteomics Core Facility . E11 . 5 dorsal cortices were dissected in cold PBS and flash-frozen in liquid nitrogen . Samples were stored in -80°C until use . 200 μl of 8M urea in 50 mM ammonium bicarbonate was added to E11 . 5 dorsal cortices . Samples were subjected to 3 rounds of probe sonication for 5s each with an energy setting of 30% . Samples were then centrifuged at 12 , 000g and 4°C for 5 minutes . All samples were run by LC/MS/MS and total ion current was used to normalize sample loading for final analysis . Samples were supplemented with 800 μl of 50 mM ammonium bicarbonate to reduce the urea concentration to 1 . 6M . Samples were reduced with 10 mM DTT at 80°C for 15 min and then alkylated at 25 mM iodoacetamide at room temperature for 30 min . Trypsin ( 1 . 7 μg ) was added to each sample and allowed to proceed for 18 hr at 37°C . Samples were then acidified with 6 μl of TFA and subjected to a C18 cleanup using the 50 mg ( 1 cc ) C18 Sep-Pak columns ( Waters ) . After elution , the samples were spun to ~50% dryness in the vacuum centrifuge , frozen , and lyophilized to dryness . Samples were randomized in their run order and QC samples were run periodically throughout the acquisition window . Samples were initially resuspended in 12 μl of 1% TFA/2% acetonitrile with 10 fmol/μl yeast alcohol dehydrogenase . To create a “QC pool” sample to assess analytical reproducibility , 3 μl of each sample was removed and pooled . Quantitative LC/MS/MS was performed on 2 μl of each sample , using a nanoAcquity UPLC system ( Waters Corp ) coupled to a Thermo QExactive Plus high resolution accurate mass tandem mass spectrometer ( Thermo ) via a nanoelectrospray ionization source . Briefly , the sample was first trapped on a Symmetry C18 300 mm × 180 mm trapping column ( 5 μl/min at 99 . 9/0 . 1v/v water/acetonitrile ) , after which the analytical separation was performed using a 1 . 7 μm Acquity BEH130 C18 75 mm × 250 mm column ( Waters Corp . ) using 90-min linear gradient of 5 to 40% acetonitrile with 0 . 1% formic acid at a flow rate of 400 nl/min with a column temperature of 55°C . Data collection on the QExactive Plus mass spectrometer was performed in a data-dependent acquisition ( DDA ) mode of acquisition with a r = 70 , 000 ( @ m/z 200 ) full MS scan from m/z 375–1600 with a target AGC value of 1e6 ions followed by 10 MS/MS scans at r-17 , 500 ( @ m/z 200 ) at a target AGC value of 5e4 ions . Following the 12 LC-MS/MS analyses , data were imported into Rosetta Elucidator v3 . 3 ( Rosetta Biosoftware , Inc ) , and all LC-MS/MS runs were aligned based on the accurate mass and retention time of detected ions ( “features” ) which contained MS/MS spectra using PeakTeller algorithm ( Elucidator ) . A mean normalization of the high confidence identified peptide features excluding the highest and lowest 10% of the identified signals ( i . e . a robust mean normalization ) was then employed . The relative peptide abundance was calculated based on area-under-the-curve ( AUC ) of aligned features across all runs . Database searching was performed within Mascot Server v2 . 5 ( Matrix Science ) and annotated using the Peptide Teller algorithm within Rosetta Elucidator at a peptide false discovery rate of 1% . Proteins representing membrane , nuclear and cytoplasmic fractions were present in the data . | The mammalian neocortex is the brain structure responsible for higher cognition , abstract thought , and language . One process critical for brain development is neurogenesis , in which neural stem cell populations generate neurons . Alterations in neurogenesis can lead to neurodevelopmental disorders affecting brain size and function , such as microcephaly , in which the brain is significantly smaller than normal . Therefore , understanding the genes and processes controlling normal brain development is of strong clinical relevance . Here we studied proteins of the RNA binding exon junction complex , which are strongly implicated in several neurodevelopmental pathologies , but whose functions in brain development remain largely unknown . Using mouse models , we find that reduced levels of any of three essential proteins of this complex results in altered embryonic neurogenesis and microcephaly . We demonstrate that mutant mice show common alterations in p53 activation , expression of ribosomal components and splice variants for RNA processing factors . Interestingly we find that genetic suppression of p53 significantly rescues microcephaly in mutant mice . Given that patients harboring mutations in exon junction complex components present with neurodevelopmental deficits , our findings highlight molecular pathways which could underlie disease pathogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"neurogenesis",
"microcephaly",
"neuroscience",
"animal",
"models",
"developmental",
"biology",
"model",
"organisms",
"genome",
"analysis",
"morphogenesis",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"developmental",
"neuroscience",
"genomics",
"animal",
"cells",
"birth",
"defects",
"mouse",
"models",
"proteomics",
"congenital",
"disorders",
"ribosomes",
"biochemistry",
"cellular",
"neuroscience",
"cell",
"biology",
"heredity",
"neurons",
"genetics",
"transcriptome",
"analysis",
"biology",
"and",
"life",
"sciences",
"haploinsufficiency",
"cellular",
"types",
"computational",
"biology"
] | 2016 | Haploinsufficiency for Core Exon Junction Complex Components Disrupts Embryonic Neurogenesis and Causes p53-Mediated Microcephaly |
Patterns of disease co-occurrence that deviate from statistical independence may represent important constraints on biological mechanism , which sometimes can be explained by shared genetics . In this work we study the relationship between disease co-occurrence and commonly shared genetic architecture of disease . Records of pairs of diseases were combined from two different electronic medical systems ( Columbia , Stanford ) , and compared to a large database of published disease-associated genetic variants ( VARIMED ) ; data on 35 disorders were available across all three sources , which include medical records for over 1 . 2 million patients and variants from over 17 , 000 publications . Based on the sources in which they appeared , disease pairs were categorized as having predominant clinical , genetic , or both kinds of manifestations . Confounding effects of age on disease incidence were controlled for by only comparing diseases when they fall in the same cluster of similarly shaped incidence patterns . We find that disease pairs that are overrepresented in both electronic medical record systems and in VARIMED come from two main disease classes , autoimmune and neuropsychiatric . We furthermore identify specific genes that are shared within these disease groups .
When two diseases occur together in the same individuals more or less often than would be expected by chance , this may signal the operation of important biological processes . Pairs of diseases occurring more than expected are called synergistic; such interactions are familiar from clinical practice when the occurrence of a disease may raise the risk of a second disease . Pairs occurring less than expected are called protective; these interactions , sometimes called “inverse comorbidities , ” are less common , but intriguing . Disease pairs which consistently diverge from independence in either direction may provide clues towards identifying core genetic , pathway , physiological , or environmental constraints that alter disease risk and represent an important starting point for elaborating a mechanistic understanding of disease and for locating possible drug targets . Because discovery of disease patterns has been haphazard , it is attractive to systematically search for these patterns across a wide range of diseases , without adhering to prior conceptions of disease class , associated features , or expected comorbidities . In this work , we integrate clinical and genomic data across diseases to systematically assess their co-occurrence . Consistent co-occurrence and conditional dependence in disease phenotypes arises from multiple , non-exclusive , factors: ( 1 ) shared genetics , including causal effects of single genes and effects of neighboring genes in linkage disequilibrium , ( 2 ) shared environmental exposures , ( 3 ) complex interactions in which a phenotype enhances or moderates the risk of another , ( 4 ) ascertainment , selection , or referral bias , ( 5 ) artifacts of the diagnostic system , where two putatively separate diseases are linked via large overlap of shared features , and ( 6 ) random variation . Untangling these factors requires use and integration of both phenotypic and genetic data . Historically , non-independent phenotype associations are noticed in an opportunistic way when the effect size is large , and otherwise they are detected more accurately through observational studies and meta-analyses [1] , or via comprehensive epidemiologic surveys . However , such studies and surveys are expensive to conduct and therefore often do not methodically examine disease combinations . In contrast , electronic medical records ( EMR ) represent a source of coded medical data that is typically large and , because these records are routinely collected to support clinical and administrative operations , the marginal cost to researchers is small; EMR data may therefore facilitate systematic comparison of disease co-occurrence [2] . Complementary information about disease relationships can be drawn from genomic studies . In particular , VARiants Informing MEDicine , or VARIMED [3] , is a hand-curated database of published disease-associated ( primarily common ) genetic variants . Although it is limited to known genetic variants , it is large and provides an opportunity for detecting the overlapping and shared genetic bases of diseases . We combine EMR data with genetic data to compare and contrast disease co-occurrence patterns , systematically comparing statistically significant disease comorbidity patterns in EMR data with disease pairs having statistically significant genetic overlap in VARIMED , and characterize the pairs by the predominant influence as ( 1 ) clinical and genetic if they both co-occur in the clinical data and share a significant genetic component , ( 2 ) clinical without genetic if they co-occur only in the clinical records , or ( 3 ) genetic without clinically observable effect if we find only a significant genetic overlap without a corresponding EMR result . There are several important assumptions to consider here including the penetrance and causality of the genetic relationships that we examine as well as interactions between the genetics and the environment . Furthermore , EMR data are prone to selection and ascertainment bias , and errors from inaccuracies in chart coding . The lifetime of the EMR induces an observation window on the patients represented there , underrecording data from patients for disease pairs with widely separated ages of disease onset , and generating false inverse comorbidities . In order to avoid the confounding effect of age on the pair occurrence counts , we introduce a method for clustering diseases through similarity of their incidence pattern by age . Other researchers have explored similar ideas . Patterns have been detected using linked administrative and clinical databases . Goldacre and colleagues [4] used data from the Oxford Record Linkage Study to find disease associations , such as an expected association between schizophrenia and lung cancer , and a protective association between schizophrenia and rheumatoid arthritis . A later study using similar data [5] found inverse associations between Parkinson’s disease and several kinds of cancer . Rzhetsky et al . [2] developed a mathematical model of ICD9-coded data from a single EMR to infer genetic overlap . Using genomics data from early GWAS studies , Sirota et al . [6] used summary data to define a signed genetic variation score and cluster autoimmune disorders . Jung et al . [7] applied a similar method to studying autoimmune disorders when paired with autism . Li et al . [8] used data from several EMRs and from VARIMED to identify the genetic architecture of novel risk factor-disease associations . Ibáñez et al . [9] compared gene expression profiles for previously identified inversely comorbid neuropsychiatric/cancer disease pairs , and found corresponding up- and down-regulation patterns . Melamed et al . [10] used data from a large database of insurance claims in combination with known genetic associations for Mendelian disorders to identify cancer driver genes . Glicksberg et al . [11] compared the overlap in disease pairs using EMR data and a database of genetic variants , retaining those pairs where both diseases appeared together in PubMed articles . In this paper , we present a framework for integrating clinical and molecular data to study disease co-occurrence . Because disease risk varies with patient age , and because the co-occurrence of disease is therefore confounded by age , we introduce a method to define age-specific disease clusters and carry out pairwise comparisons of disease co-occurrence . We explicitly model disease pair under and overrepresentation . To reduce bias , we conduct the analysis in two independent clinical databases , and require statistically significant deviation from independence in both . We identify a highly significant group of autoimmune disorders , a set of diseases with known environmental triggers , and some results which question the clinical manifestation of previously described disease associations .
In this section we report the results that are significant in both EMRs ( Columbia and Stanford ) and in VARIMED . We refer to these as “clinical and genetic . ” For disease pairs significant in Columbia , Stanford , and in VARIMED , none were protective , and five were synergistic . The results , which fall into two classes , autoimmune and neuropsychiatric , are shown in Table 4 . Information about the genes and gene overlap for the five overrepresented pairs appears in Table 5 . Prior work has found considerable genetic sharing between many autoimmune diseases [6] , [12]; specific results include , ankylosing spondylitis and psoriasis [13] . The association between ankylosing spondylitis and lupus has been reported , but is extremely rare [14] . Rheumatoid arthritis and secondary Sjogren’s syndrome have a well-known association . The other two results are previously identified associations between neuropsychiatric disorders; bipolar disorder and schizophrenia [15] , and bipolar disorder and depression , although there may also be diagnostic overlap , as depression and bipolar disorder can be confused clinically . We furthermore identify specific genes which are common to these two groups ( Table 5 ) . In the autoimmune subgroup those include well known associations in the HLA region such as HLA-DRA , HLA-E[16] , interleukin receptors ( IL13 , IL23R and IL2RA ) [17 , 18] , [19] , BTNL2[20] and MICA[21] . Interleukins are any of a class of glycoproteins produced by leukocytes for regulating immune responses . While these genes have been previously associated with autoimmune diseases , they provide an interesting opportunity to explore shared therapeutic targets and diagnostic markers across these phenotypes . In the neuropsychiatric subgroup some genes that are of interest include ANK3 , CACNA1C , CDH13 , ITIH4 and PDE7B . Ankyrins are a family of proteins that are believed to link the integral membrane proteins and play key roles in activities such as cell motility , activation , proliferation , contact , and the maintenance of specialized membrane domains . Ankyrin 3 is an immunologically distinct gene product from ankyrins 1 and 2 , and was originally found at the axonal initial segment and nodes of Ranvier of neurons in the central and peripheral nervous systems . CACNA1C is a voltage-dependent calcium channel and has been previously linked to several neurodegenerative diseases [22] , [23] . CACNA1C is also an associated gene of the one of the most highly significant SNPs for both bipolar disorder and schizophrenia in a cross-disorder genome wide analysis [15] . Cadhedrin , CDH13 , is a known ADHD-susceptibility gene that has been investigated in other neuropsychiatric disorders [24] , [25] , [26] . Some of the other shared genes such as ITIH4 , inter-alpha-trypsin inhibitor heavy chain family , member 4 , and PDE7B , phosphodiesterase 7B do not have clearly known links to the neuropsychiatric phenotypes and might be interesting to explore further . In this section we report disease pairs that are significant in both EMRs , but not significant in VARIMED ( “clinical without observed genetic effect” ) . One protective interaction was found: alcoholism and goiter . This pair in the Columbia dataset has an observed/expected ratio of 0 . 501 ( p < 1 . 55 × 10−22 ) ; in the Stanford dataset , 0 . 297 ( p < 6 . 16 × 10−61 ) . Table 6 shows the 23 overrepresented interactions . As expected , disorders with clear environmental triggers are apparent in both lists: alcohol , injection drug use ( HIV , hepatitis B and C ) , and diet ( diabetes type 2 , and gout ) . Of the protective pairs , alcoholism and goiter have been previously noted to be underrepresented [27] . Most of the detected synergistic interactions are well-known . These include: alcoholism and bipolar disorder , alcoholism and depression , alcoholism and schizophrenia , depression and schizophrenia , and the alcoholism and injection drug-associated pairs . Of those less familiar , references are provided here: depression and migraine [28] , migraine and lupus [29] , cardiomyopathy and diabetes [30] , aortic aneurysm and cardiomyopathy [31] , diabetes type 2 and gout [32] , attention deficit and autism [33] . The association between alzheimer’s and parkinsonism is very likely due to diagnostic overlap , given known differences in mechanism but difficulties in the clinical diagnosis of dementia subtypes . Lack of clear genetic signal for all these pairs does not completely rule out any genetic connection , as some disorders may have not yet been subject to scrutiny through GWAS studies , or have only modest effect sizes not reaching statistical significance . In this section we report disease pairs that have significant overlap of genetic variants in VARIMED , but are not significant in both EMRs . There are 17 such pairs , shown in Table 7; when the disease pair was significant in one EMR , that was recorded in the “EMR” column . The group of “genetic without observed clinical effect” are those which have significant genetic overlap in VARIMED , but not in both EMRs . Nearly all are autoimmune disorders , and may represent pairs with sharing detected at the level of genes that do not produce pathway interactions leading to disease phenotypes , or rare interactions that do not achieve statistical significance .
In this paper we present a method to identify statistically significant disease pairs which display significant comorbidity in two EMRs and share common genetic background in a large database of disease-associated variants; we explicitly model the under and overrepresentation of disease pairs , and control for the confounding effects of age on disease incidence by only comparing diseases when they fall in the same cluster of similarly-shaped incidence patterns . The method is fast , easy to interpret , and can be extended in a straightforward manner to other EMRs , data from national health systems , and large insurance databases . Our primary aim is to identify disease pairs which might share a common mechanism or treatment option for further exploration and research . We link disease pairs that are under or overrepresented in EMR data to statistically significant overlapping genes sets for the same pairs . The genetic variants are known to have phenotypic effects , while EMRs capture a broad collection of diseases states that are severe enough to require diagnosis and treatment , and represent a constellation of genetic predispositions , environmental influences , and social and economic factors that affect when diseases are detected . Many of the predisposing factors in EMRs are not measured , but we can find pairs that have known genetic associations and also find pairs that do not . As always for candidate generation or prioritization methods , the question arises of how to validate novel results , given that validating experiments have not yet been conducted . By contrasting results in two EMRs and in a database of genetic variants , we have reduced the chance that the same biases are operating across all data sources . There are several limitations of our approach which should be recognized . Our method only compares diseases when they fall in the same cluster . This is a simple , but conservative , match on age patterns , and should enrich results for true positives at the expense of missing other true positives that would only be found in cross-cluster comparisons . For example , because autism and Alzheimer’s disease would fall in different age-incidence clusters and would not be compared , possible interactions between those disorders would not be detected . VARIMED , while large , contains only published results , reflecting investigators’ choices of important areas of study , including , as we found , autoimmune disorders and neuropsychiatric disorders . Also , VARIMED focuses primarily on common variation as most genetic association has been based on genotype-based GWAS . VARIMED ( and other databases ) are not randomly sampled from the space of biological phenomena , and the absence of a genetic variant may only mean that such have not yet been investigated . It is likely that our method will fail in such circumstances to identify comorbid pairs using the conjunction of data from EMRs and from VARIMED , which is a source of bias . In addition , we link to EMR records on the basis of a straightforward , but necessarily imprecise , mapping through a disease name . We restrict the genetic analysis to the genes and do not consider the allele-specific relationships ( risk-enhancing or risk-moderating ) . Although both gender and ethnicity are known to be important covariates for the prevalence of disease , because the available Columbia data were not stratified by gender or ethnicity , neither were used in this study . This would be particularly important for autoimmune disorders with their known gender dependence; combining the genders for analysis , as we had to do , may have diluted statistical signal , and would explain the appearance of results in Table 7 . Finally , the set of diseases examined was restricted to the 161 in the original Rzhetsky study , and further restricted by the limited overlap with VARIMED; although drawing from both common and rare diseases , the set is small compared to the full range coded by ICD9 . Prior studies have found multiple protective interactions between CNS disorders and cancers [34] but , unfortunately , few cancers were in the list of disorders analyzed here . In spite of these limitations , we hope this study can serve as a proof of principle for integrating EMR and genetics data to uncover relationships between diseases . Using larger data sets , and incorporating important covariates and the direction of allele-specific risk would important validating extensions of the current work . Furthermore , text mining of EMR clinical notes and other databases of environmental exposures could represent an opportunity for identifying non-genetic causes of diseases . In conclusion , we have presented a method integrating clinical EMR and genetics data in order to elucidate disease comorbidity . We identify a set of disease pairs which deviate from the independence assumption in their co-occurrence in two different EMR systems . By integrating the clinical observations with genetics , we are further able to categorize which of the disease pairs might be explained by the shared genetics and which might have more of an environmental component .
Our validation data set used patient records from Stanford’s electronic medical record system , STRIDE ( Stanford Translational Research Integrated Database Environment ) . The data request was judged to be exempt from human subject concerns by the Stanford Institutional Review Board , and was also approved by its Data Privacy Office . The Stanford data were retrieved June 6 , 2013 . Encounter records contained a masked patient identifier , current age , gender , ethnicity , icd9 , and age at visit . ( Gender was not used because gender information was not generally available for the Columbia data . Ethnicity was not used for the same reason . ) Because of small numbers of very old patients , ages were censored at 90 years for privacy reasons by STRIDE staff prior to our use . For the electronic medical record data , the discovery data set comes from the composite data for the Columbia EMR , published as an online appendix of [2] , which lists counts of diseases and disease pairs for a total of 161 disorders . As described in the original article , “We selected disorders that represent a broad spectrum of maladies , from common to rare , affecting diverse physiological systems , yet we also placed special emphasis on neurological phenotypes . ” The total number of patients was 1 , 478 , 976 , however because these records include data on healthy hospital employees , the total was lowered by 500 , 000 as described in their Appendix 2 , p 19 . Disease count data were extracted from their Appendix 3 , and disease-pair count data were extracted from Supplemental Information Data Set 1 . Separately , the mapping from ICD9 codes to disease names was taken from their Appendix 3; a small number of mapping errors were corrected by hand . For validation , patient-level data were retrieved from STRIDE ( Stanford Translational Research Integrated Database Environment ) [35] for the same 161 diseases to allow for direct comparison with the Columbia data . The raw data contained 1 , 057 , 132 records for 397 , 474 patients . We focus our analysis on data starting in the year 2008 , which was the year of comprehensive EMR rollout . When there were fewer than 50 patients with a disease , that disease was judged too rare to contribute to the incidence frequencies in a meaningful way , and was removed . This left data for 277 , 290 patients . Also , disease pairs were removed if any of the cells in the 2 × 2 table had observed or expected values less than 5 . The EMR records were aggregrated and processed to retain the earliest occurrence of each ICD9 code for each patient , which were then consolidated using the ICD9-to-disease mapping from [2] to produce a table of patient counts for each disease name . A similar procedure was used to count disease pairs . Biases arise from EMR data not being a random sample of diseases in the population . For example , autism and Alzheimer’s disease have very different incidence patterns . ( See Fig 5 . ) It would be unlikely for a patient to have this disease pair in their records , even if they were ultimately afflicted by both disorders , because young patients at risk for autism would not also be at risk for Alzheimer’s until many years in the future , and those at risk for Alzheimer’s would have been at risk for autism in an era when the EMR did not exist , even if autism had then been a clearly defined syndrome . This will lead to systematic undercounting of these and similar disorder pairs . To control for the confounding effects of age , disease pairs were analyzed only when both diseases could be put in the same age-incidence cluster . Clusters were formed so as to be as large as possible ( to maximize the number of subsequent disease comparisons ) , while simultaneously imposing within-cluster homogeneity , so that each cluster had similar age-incidence patterns . The age-incidence clustering used from patient-level data for Stanford; corresponding details for the Columbia data were not available . Each disease was represented as a 91-dimensional vector containing counts of the number of patients whose earliest onset of that disease occurred for each of the ages 0 years through 90 years . For normalization , each vector was divided by its length to produce unit vectors . Hierarchical clustering with Ward’s method for linkage was chosen to produce clusters that were compact and of similar size . Cluster size was determined in a data-driven manner by systemically searching through possible clustering methods and cluster scoring measures . The methods and measures were taken from those provided by the R package COMMUNAL and are listed in the Supplemental Material . The cluster measures ( also known as cluster indices ) provide scores for each method and each cluster size . The measures were combined into a composite score by standardizing each measure ( zero mean , unit variance ) and then averaging . All measures were converted to have the same sense , so that larger values were associated with more desirable clusters . Any measure with a monotonic function ( either increasing or decreasing ) of cluster size for all methods and measures was removed because such a measure would be minimized or maximized at the extremes of the search range for cluster size , and thus not be responsive to patterns in the data . Pairs of diseases that showed significant comorbidity pairs were identified in the Columbia data and verified in the Stanford data , so all pairs reported here were statistically significant in both . In addition , only pairs that were underrepresented in both EMRs or overrepresented in both EMRs were retained , ensuring consistent directionality; discordant pairs were not analyzed . Statistical significance was computed using the Fisher exact test . Bonferroni correction was applied using the number of diseases in each cluster . The conventional level of significance , 0 . 05 , was used for all tests . Genetic associations came from VARIMED , a hand-curated database of published phenotype-associated genetic variants [3] . As of May , 2015 , this database contained variants from 17 , 088 publications , with 466 , 890 SNPs associated with 2 , 992 diseases or traits . SNPs were mapped to genes using the dbSNP annotation database . Sometimes , a variant maps to more than one gene . In such cases , we use colons to separate the genes in a single group; this notation is used in Table 5 . Phenotype descriptions were mapped by hand to the set of 161 disease names used for the Columbia and Stanford data . There were 35 diseases that appeared in Columbia , Stanford , and in VARIMED ( Table 3 ) . Because VARIMED is proprietary , the relevant subset of 35 diseases , with associated genes , chromosome number , and PubMed ID of the source of each association were extracted and used for the analysis we report here . This dataset is included in the Supplement . Genetic variants that were significantly associated with each phenotype of interest were obtained from VARIMED and mapped to gene names . In this study , we used significant disease-SNP associations ( p < 10−6 ) with known risk alleles and published odds ratios . The number of genes associated with each of the 35 diseases of interest are shown in Table 3 . We furthermore focus our analysis on the gene level , specifically calculating enrichment of the number of overlapping genes between two phenotypes of interest . We report the number of genes shared by the disease pair if the overlap was determined as significant by the Fisher exact test using Bonferroni correction for the number of tests . A network diagram ( Fig 3 ) showing the structure of the disease pairs and their clusterings was created using the Cytoscape software tool . In the network diagram , a node represents a disease . Two nodes are connected if that disease pair is statistically significant in the EMR data and appears in the same age-incidence cluster . The size of a node represents the frequency of that disease in the larger EMR ( Columbia ) . The edge width represents the effect size for that pair ( observed number divided by expected number ) . The node color indicates cluster membership , using the same colors as in Fig 2 . | Diseases do not always occur together at random , and patterns of disease association may reflect important biological constraints on disease mechanism . When a disease pair occurs more ( or less ) often than would be expected by chance given the frequencies of each disease separately , that may signal the presence of shared causal factors . These shared factors may be genetic , environmental , or interactions of the two . To characterize the kinds of possible disease causes , we compared data from electronic medical records , which record disease manifestations , with information about genetic variants of disease . In particular , we find pairs of diseases that occur more or less often than expected by chance for patients in two electronic medical record systems , and compare them to disease pairs sharing a statistically significant number of genes in a database of disease-associated genetic variants . Overrepresented pairs appearing in both source types come from two main disease classes , autoimmune and neuropsychiatric . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"bipolar",
"disorder",
"social",
"sciences",
"autism",
"developmental",
"psychology",
"neuroscience",
"diabetes",
"mellitus",
"endocrine",
"disorders",
"autism",
"spectrum",
"disorder",
"mood",
"disorders",
"alcoholism",
"public",
"and",
"occupational",
"health",
"developmental",
"neuroscience",
"neurodevelopmental",
"disorders",
"addiction",
"endocrinology",
"schizophrenia",
"mental",
"health",
"and",
"psychiatry",
"metabolic",
"disorders",
"clinical",
"genetics",
"psychology",
"neurology",
"genetics",
"biology",
"and",
"life",
"sciences",
"depression",
"substance-related",
"disorders",
"genetics",
"of",
"disease"
] | 2016 | Constraints on Biological Mechanism from Disease Comorbidity Using Electronic Medical Records and Database of Genetic Variants |
Detection of Mycobacterium leprae in slit skin smear ( SSS ) is a gold standard technique for the leprosy diagnosis . Over recent years , molecular diagnosis by using PCR has been increasingly used as an alternative for its diagnosis due to its higher sensitivity . This study was carried out for comparative evaluation of PCR and SSS microscopy in a cohort of new leprosy cases diagnosed in B . P . Koirala Institute of health Sciences , Dharan , Nepal . In this prospective crossectional study , 50 new clinically diagnosed cases of leprosy were included . DNA was extracted from SSS and PCR was carried out to amplify 129 bp sequence of M . leprae repetitive element . Sensitivity of SSS and PCR was 18% and 72% respectively . Improvement of 54% case detection by PCR clearly showed its advantage over SSS . Furthermore , PCR could confirm the leprosy diagnosis in 66% of AFB negative cases indicating its superiority over SSS . In the paucibacillary ( PB ) patients , whose BI was zero; sensitivity of PCR was 44% , whereas it was 78% in the multibacillary patients . Our study showed PCR to be more sensitive than SSS microscopy in diagnosing leprosy . Moreover , it explored the characteristic feature of PCR which detected higher level of early stage ( PB ) cases tested negative by SSS . Being an expensive technique , PCR may not be feasible in all the cases , however , it would be useful in diagnosis of early cases of leprosy as opposed to SSS .
Leprosy is a chronic infectious disease which mainly affects the skin , nasal mucosa , and peripheral nerve . It is caused by Mycobacterium leprae , an acid-fast bacillus which is transmitted via droplets from the nose and mouth during close contacts with untreated cases [1] . It eventually leads to disability , disfiguration and social stigma for the rest of patient’s life if untreated . Therefore , early detection of M . leprae is the key element to timely identification and treatment of patients before nerve involvement occurs . Diagnosis of leprosy has traditionally been based on clinical examinations and skin smears [2] . Microscopy has advantage of being easily available at peripheral and referral centers but as its detection limit is 104 bacilli/ml , it suffers from low sensitivity [3] . M . leprae does not grow in vitro . In contrary to microscopy , DNA based PCR have shown superior performance to detect bacilli in clinical samples [4–8] . However , their application is not wide in resources poor countries . Elimination of leprosy as a public health problem at the global level was achieved by 2000; elimination was defined as a reduction in prevalence to <1 case per 10 , 000 population [9] . Nepal has achieved the elimination goal in December 2009 , but then the incidence of new case & prevalence rate has increased from 0 . 77 to 0 . 79 , 0 . 84 , 0 . 82 and 0 . 83 respectively during fiscal year ( 2009/10 ) to ( 2013/14 ) [10] . Despite achieving the official leprosy elimination goal , it is still considered as a smoldering problem as several cases are under reported to health care facilities . It shows that post elimination phase is more challenging as to maintain the low number of case load which is entirely dependent upon the precise diagnostic tools . In this context , present study was conducted to compare the performance of PCR with the existing slit skin smear for detection of clinically diagnosed cases of leprosy .
This is a prospective crossectional study . Consecutive new clinically diagnosed cases of leprosy who presented to department of dermatology and venereology during April 2012 to March 2013 were considered for enrollment in the study . Laboratory investigations were carried out in the department of microbiology . Those patients who did not give consent and those patient under treatment or having past history of treatment received were not included . A case of leprosy was defined by the presence of any one of the following cardinal signs [11] . Cases were clinically categorized according to Ridley-Jopling classification ( RJ ) [12] into: Tuberculoid ( TT ) , Borderline tuberculoid ( BT ) , Borderline ( BB ) , Borderline lepromatous ( BL ) and Lepromatous ( LL ) . Likewise the cases were also classified according to WHO classification [13] system as Paucibacillary ( PB ) and Multibacillary ( MB ) . Ethical clearance to undertake the study was obtained from Institutional review board of B . P . Koirala Institute of Health Sciences ( BPKIHS ) , Dharan , Nepal . Written informed consent was obtained from all the adult patients . Previous study[20] reported sensitivity of PCR ( index test ) and SSS ( reference test ) of 86% and 60% respectively . With power of 90% with α error 5% ( one sided ) , fifty patients were required in this study . Data were statistically described in terms of range , mean ± standard deviation ( SD ) , frequency ( number of cases ) , relative frequency ( percentages ) , and confidence interval when appropriate . The statistical significance of the differences in sensitivities between PCR and SSS microscopy were assessed by means of Chi square test , Fisher’s exact test and kappa test . SPSS ( version 15; Chicago , IL ) was used for statistical analysis . In addition , the ROC curve was analyzed for PCR diagnosis with RJ Clinical Types by R version 3 . 0 . 3 ( www . r-project . org ) [21] applying the package pROC [22] .
The study included 50 patients who met the criteria of case definition of leprosy and did not have history of past treatment for leprosy . Consecutive SSS microscopy was performed from 50 patients within 1 year and then , PCR was performed on all 50 SSS . The flow of participants through the study is shown in Fig 2 . There were 27 ( 54% ) male and 23 ( 46% ) female . Mean age of patient was 42 . 12 ± 17 . 11 years ( range: 10–75 years; median: 44 . 50 years ) . Majority ( 66% ) of the lesions had well defined margin . Nerve involvement was found in 44 ( 88% ) of the patients on peripheral nerve examination with ulnar nerve being commnest . Eye involvement was present in 9 ( 18% ) . Out of 50 patients , microscopy was positive in 9 [18%; 95%CI , ( 9 . 54–31 . 02 ) ]cases . In contrast , PCR detected M . leprae in 36 [72%; 95%CI , ( 58 . 24–82 . 62 ) ]cases ( Fig 3 ) . There was a 129 bp DNA band in agarose gel electrophoresis , indicating presence of M . leprae as shown in Fig 4 . A total of 9 ( 18% ) cases tested positive by both diagnostic tools i . e . modified AFB staining and PCR . There was no case in which microscopy was positive and PCR was negative ( Table 1 ) . PCR confirmed the diagnosis in 27 ( 66% ) out of 41 skin smears which were AFB negative . Further comparative analysis of positivity of AFB in SSS and PCR to the various subgroups as per RJ clinical types and WHO clinical types are depicted in Tables 2 and 3 respectively . Receiver Operating Characteristic ( ROC ) curve analysis was performed for PCR with RJ classification ( Fig 5 ) . The diagnostic accuracy was 0 . 846 ( area under curve; AUC = 0 . 846 , 95% CI = 0 . 7698–0 . 9225 ) .
SSS being cheap and minimally invasive , are the main tool for diagnosis of leprosy in the developing countries where most cases are newly detected [9] . However , this conventional technique has the disadvantage of being less sensitive [3] . There is need for a more sensitive diagnostic tool for early diagnosis of leprosy cases to prevent the deformities and disabilities . Several studies have reported successes to detect M . leprae by PCR in SSS showing its clear advantage over SSS microcopy [8 , 16] . Therefore , the present study was carried on 50 new patients comparing the routine conventional SSS microscopy with PCR on SSS in the leprosy diagnosis . Our study demonstrated a positive yield in 18% of all cases of clinically diagnosed leprosy by SSS microscopy , while 72% were positive by PCR . As the 95% confidence interval of sensitivity of microscopy and PCR do not overlap in Fig 3 , therefore , this result showed the difference in the diagnostic efficacy of these two diagnostic tool to be statistically significant [23] . Moreover , the ROC curve of PCR indicates it as a good diagnostic tool since the AUC is 0 . 846 as shown in Fig 5 [24] . Hence , this study shows the diagnostic efficacy of PCR is more efficient than microscopy in diagnosing leprosy . The low sensitivity of SSS microcopy can be explained by individual observer variation and low load of bacteria in SSS . An improvement of 54% in the case detection rate as compared to the results of SSS alone definitely mirrors out the better diagnostic efficiency of PCR over microscopy . Furthermore , the PCR sensitivity could have been greater , but xylene and other chemical components used during DNA extraction may have inhibited the PCR . Our findings of PCR with overall 72% yield was comparable to that of Kamal et al [25] with 72% overall yield with in situ PCR on SSS . However , this rate was still lower as compared to the studies conducted in Thailand [8] and Kolkata [20] in which PCR positivity was noted to be 87% and 82 . 3% respectively in the biopsy samples . This difference can be explained by the low level of DNA in SSS due to less amount of tissue as compared to biopsy . It is interesting to note that our results were still better with 12% more than the Dayal et al [26] , study in which in situ PCR was used in skin biopsy resulting into positivity of 60% . Difference in the nature of the PCR primer ( 36kDa antigen gene ) as compared to that of ours ( RLEP LP1/LP2 ) may be the possible reason for the discrepancy in the findings . Compared with 36-kDa antigen gene primers , the RLEP primers were found to be 1000-fold sensitive in detection of M . leprae DNA in a study performed in London [19] . In the present study , 18% of the total cases were SSS positive with 100% in BB , 67% in LL and 14% in BL subgroups . Several studies have reported similar rate sensitivities . In the study by Kamal et al [25] , 20% cases were skin smear positive with 44% positivity in BL/LL variety . Similar rate of positivity of SSS was observed by Dayal et al [26] with only 10% positivity out of total cases and that of BL variety . All the PB ( BT and TT ) cases in the present study tested negative in the SSS , a finding corresponding the results reported by both of the above cited studies in which skin smears were negative in the indeterminate and BT variety . The scarcity of AFB in skin tissues in the early stages ( PB ) usually causes difficulty in microscopic diagnosis . Our study found the PCR positivity rate of 66% in AFB negative cases which was similar to other studies findings of 65% [20] and 72 . 7% [26] in the cases which were AFB negative . This highlights the superiority of PCR over SSS which could confirm the leprosy diagnosis in more than half of the cases missed by SSS . Furthermore , our study found the sensitivity of PCR on AFB- positive cases to be 100% indicating no false positive on SSS findings . The number of PCR-positive cases was detected to be higher in MB than in PB patients in our study . As many as 78% of the MB cases were PCR positive , while this rate was only 44% in the PB cases . This result was expected as MB leprosy has a higher bacterial load than PB leprosy . Further , our results are better than the results of another study conducted at Thailand [8] using PCR on SSS . In the latter study , 41 . 9% positivity was observed in the MB cases and 18 . 2% in PB cases . But in the study conducted in Vietnam [27] , 100% PCR positivity was recorded on the SSS samples from the MB patients , a finding superior to ours . Alternatively , while the results of our PCR test sensitivity was convincing , but this assay was applied only on SSS and not on other clinical specimens which is the limitation of this study . In summary , our study clearly reveals that PCR is more sensitive than conventional SSS microscopy in diagnosing leprosy . Moreover , this study explored the characteristic feature of PCR which detected higher level of early stage ( PB ) cases tested negative by SSS . Being an expensive technique , PCR may not be feasible in all the cases , however , it would be useful in diagnosis of early cases of leprosy as opposed to SSS . Besides , PCR is only accessible in few teaching hospitals and referral centers in resource poor country like Nepal . Therefore , early and clinically indeterminate cases can be referred to the referral center for PCR to increase the case detection of leprosy . | Although leprosy has been eliminated at the national level , but region wise eastern Nepal has still the higher number of leprosy cases . Early diagnosis and then timely treatment is very crucial to prevent disabilities due to leprosy . Slit-skin smear ( SSS ) , the routine diagnostic method was compared with the molecular method polymerase chain reaction ( PCR ) for the detection of new clinically diagnosed cases at BPKIHS . PCR was found to be much more superior than SSS in its diagnostic efficacy detecting more than half number of cases missed by SSS . Moreover our finding of PCR in detection of early stages was nearly 50% where SSS are mostly negative on such spectrum . Therefore , this costly but sensitive diagnostic tool PCR was found to be a better option in detection of early cases of leprosy in comparison to SSS . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"mycobacterium",
"leprae",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"nervous",
"system",
"tropical",
"diseases",
"geographical",
"locations",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"bacteria",
"extraction",
"techniques",
"nepal",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"artificial",
"gene",
"amplification",
"and",
"extension",
"lesions",
"actinobacteria",
"molecular",
"biology",
"people",
"and",
"places",
"dna",
"extraction",
"diagnostic",
"medicine",
"anatomy",
"nerves",
"asia",
"polymerase",
"chain",
"reaction",
"leprosy",
"biology",
"and",
"life",
"sciences",
"organisms"
] | 2016 | Evaluation of Polymerase Chain Reaction (PCR) with Slit Skin Smear Examination (SSS) to Confirm Clinical Diagnosis of Leprosy in Eastern Nepal |
Complexity and heterogeneity are intrinsic to neurobiological systems , manifest in every process , at every scale , and are inextricably linked to the systems’ emergent collective behaviours and function . However , the majority of studies addressing the dynamics and computational properties of biologically inspired cortical microcircuits tend to assume ( often for the sake of analytical tractability ) a great degree of homogeneity in both neuronal and synaptic/connectivity parameters . While simplification and reductionism are necessary to understand the brain’s functional principles , disregarding the existence of the multiple heterogeneities in the cortical composition , which may be at the core of its computational proficiency , will inevitably fail to account for important phenomena and limit the scope and generalizability of cortical models . We address these issues by studying the individual and composite functional roles of heterogeneities in neuronal , synaptic and structural properties in a biophysically plausible layer 2/3 microcircuit model , built and constrained by multiple sources of empirical data . This approach was made possible by the emergence of large-scale , well curated databases , as well as the substantial improvements in experimental methodologies achieved over the last few years . Our results show that variability in single neuron parameters is the dominant source of functional specialization , leading to highly proficient microcircuits with much higher computational power than their homogeneous counterparts . We further show that fully heterogeneous circuits , which are closest to the biophysical reality , owe their response properties to the differential contribution of different sources of heterogeneity .
On a macroscopic level , hierarchical modularity is easily identifiable as a parsimonious design principle underlying various structural and functional aspects of cortical organization [19–23] . Different anatomophysiological [24 , 25] , genetic / biochemical [26–30] or functional [31–33] criteria give rise to slightly different modular parcellations but , in combination , these criteria reveal the relevant ‘building blocks’ , the most important features whose variations and recombinations give rise to the complexity and diversity of the cortical tissue [34] . For convenience , these features can be coarsely ( and tentatively ) grouped into neuronal , synaptic and structural components ( see also [13] ) . Neuronal features refer to the different cell classes and their laminar and regional distributions [35] along with their characteristic electrophysiological and biochemical diversity [36 , 37] . Synaptic components refer to a molecular default organization characterized by variations in the differential expression and transcription of genes involved in synaptic transmission [26 , 28] , which is reflected , for example , in regional receptor architectonics [38–40] . Structural aspects include variations in cortical thickness and laminar depth [41] along with neuronal and synaptic density [42 , 43] and input-output ( both local and long-range ) connectivity patterns [44 , 45] . In combination , these features highlight default organizational principles whose variations across the cortical sheet are likely to contribute to the corresponding functional specializations . Based either on morphological , electrophysiological or biochemical features ( or , preferably , a combination thereof ) , several different classes of neurons can be identified throughout the neocortex ( see e . g . [35–37 , 46–48] ) . Apart from pronounced regional and laminar differences in the types of neurons that make up the cortex and their relative spatial distributions , every microcircuit in every cortical column is composed of diverse neuron types , with heterogeneous properties and heterogeneous behaviour . Electrochemical communication between these diverse neuronal classes is an intricate , dynamic and very complex process involving a multitude of nested inter- and intracellular signalling networks [49–51] . Their functional range spans multiple spatial and temporal scales [52–54] and has , arguably , the most critical role in modulating microcircuit dynamics and information processing within and across neuronal populations [2 , 3 , 55] . The specificities of receptor composition and kinetics underlie the substantial diversity observed in the elicited post-synaptic potentials [56 , 57] across different synapse and neuronal types ( see , e . g . [58–62] ) . This occurs because the receptors mediating these events have distinct biochemical and physiological properties depending on the type of neuron they are expressed in and , naturally , the type of neurotransmitter they are responsive to . These varying properties have known and non-negligible implications in the characteristic kinetics of synaptic transmission events occurring between different neurons [63] and strongly constrain the circuit’s operations . Additionally , cortical microcircuits are not randomly coupled , but over-express specific connectivity motifs [64–70] , which bias and skew the network’s degree distributions [71] and/or introduce correlations among specific connections [72] , thus selectively modifying the impact of specific pre-synaptic neurons on their post-synaptic targets . Analogously to the heterogeneities in neuronal and synaptic properties , such structural features are known to significantly impact the circuit’s properties [73–77] . As discussed above , the variations in the anatomy and physiology of a cortical microcircuit are experimentally well established and have been shown to influence computational properties . However , the absence of complete descriptions of biophysical heterogeneity and well-substantiated empirical evidence to support them ( primarily due to technical limitations ) , has forced computational studies to take a simplified approach . This has multiple additional advantages , such as greater likelihood of analytical tractability and lower overhead for the researcher in specifying the network parameters . Simplified , homogeneous models of spiking networks have proven to be a valuable tool for a theoretically grounded exploration of microcircuit dynamics , emerging from the interaction of excitatory and inhibitory populations [78–81] . The generic principles established by studying simple balanced random networks have subsequently been applied to model specific cortical microcircuits with integrated connectivity maps and realistic numbers of neurons and synapses [82] . This approach revealed that some prominent features of spontaneous and stimulus-evoked activity and its dynamic flows through a cortical column can be accounted for by the macroscopic connectivity structure , mediated by local and long-range interactions [83] . However , by focusing on emergent dynamics , these studies neglect the functional aspects and the fact that cortical interactions serve computational purposes ( but , see [84] for a study on the computational properties of the [82] microcircuit model ) . In addition , although there are good reasons for taking a minimalist approach , assuming uniformity and homogeneity on every component of the system tends to lack cogency with respect to established anatomical and physiological facts and to disregard biophysical and biochemical plausibility . Some of these limitations were circumvented by [85] , who not only accounted for detailed and empirically-informed connectivity maps , but also employed more biologically motivated models of neuronal and synaptic dynamics and placed them in an explicit functional/computational context . In line with the results obtained with the simpler microcircuit models [82 , 84] , this study demonstrated that considering realistic structural constraints is beneficial and significantly improves the computational capabilities of the circuit . Several other studies provide important steps to move away from homogeneous systems by incorporating variability ( e . g . [75 , 86 , 87] ) , but tend to do so in a relatively arbitrary manner and/or focusing on specific forms of heterogeneity while retaining homogeneity in other components ( depending on the scientific objectives of the study ) . A completely different set of priorities for modelling cortical microcircuits are espoused by [88] in the framework of the continuous efforts of the Blue Brain project [89] . The Blue Brain approach lies on the other extreme of the descriptive scale , in that it attempts to model a cortical column in full detail , explicitly accounting for the complexities of cellular composition ( based on neuronal morphology and electrophysiology ) , synaptic anatomy and physiology , as well as thalamic innervation , essentially constituting an in silico reconstruction of a cortical column ( see also [90] ) . This approach is , naturally , extremely computationally expensive and its explanatory power is limited . The model complexity at this end of the spectrum is so close to the biophysical reality that it might not lend itself to a comprehensive understanding of dissociable and important functional principles any more readily than studying the real thing does . Nonetheless , it provides valuable insights in that it carefully replicates a lot of in vivo and in vitro responses of a real cortical column , while generating a wealth of complete and comprehensive data [88 , 91] . Thus , we conclude that while simpler models are preferable , as they are generic enough to be broadly insightful and allow us to uncover general principles , we should ask the question: what is the cost of simplification ? If a model simplifies away the core computational elements of the system , our ability to account for its operations is lost . The findings discussed above indicate that heterogeneity may be critical for the mechanisms of computation; therefore models aiming at uncovering computational principles in specific biophysical systems , such as a cortical column or microcircuit , should account for these features . In this study , we attempt to bridge this descriptive gap by building microcircuit models , inspired and constrained by the composition of Layer 2/3 , that account for key heterogeneities in neuronal , synaptic and structural properties . We implement all types of heterogeneity such that they can be switched on or off , thus enabling us to systematically disentangle and evaluate the roles played by the different types of heterogeneity in the different tentative building blocks , and how they collectively interact to shape the circuit’s dynamics and information processing capacity . The choices and characteristics of the models and parameter sets used throughout this study , as well as the general microcircuit composition are constrained and inspired by multiple sources of experimental data ( see section Data-driven microcircuit model and S1 Appendix ) and account for the prevalence of different neuronal sub-types and their heterogeneous physiological and biochemical properties , the specificities of instantaneous synaptic kinetics and its inherent diversity as well as specific structural biases in cortical micro-connectivity . All models and model parameters were , as far as possible , chosen to directly match relevant experimental reports and minimize the introduction of arbitrary model parameters , in order to ensure that the effects observed are caused by realistic forms of complexity and heterogeneity and avoid imposing excessive assumptions or preconceptions on the systems studied , i . e . to “allow biology to speak for itself” . In section Data-driven microcircuit model ( complemented by the Methods section and the Supplementary Materials ) , we explain all the details of the models and model parameters used to build and constrain the microcircuit , as well as the underlying empirical observations that motivate the choices . After specifying and fixing all the relevant parameters to , as closely as possible , match multiple sources of empirical data , we study the effects of heterogeneity on population dynamics in a quiet state , where the circuit is passively excited by background noise ( section Emergent population dynamics ) and in an active state , where the circuit is directly engaged in information processing ( section Active processing and computation ) . We evaluate the circuit’s sensitivity and responsiveness , as well as its memory and processing capacity , demonstrating a clear and unambiguous role of heterogeneity in shaping the proficiency of the system by greatly increasing the space of computable functions .
In this section , we describe the process of building a complex data-driven cortical microcircuit model capturing some of the fundamental features of layer 2/3 . We specify the detailed architecture , composition and dynamics of the microcircuits explored throughout this study as well as the motivation behind all model and parameter choices . In each relevant section , we highlight the differences between the respective homogeneous and heterogeneous conditions . A summarized , tabular description of the main models is provided in S1 Table , along with a list of the primary sources of experimental data used to constrain the model parameters , provided in S1 Appendix . All the circuits analysed throughout this study are composed of N = 2500 point neurons ( roughly corresponding to the size of a layer 2/3 microcircuit in an anatomically defined cortical column; [92] ) , divided into NE = 0 . 8 N excitatory , glutamatergic neurons and NI = 0 . 2 N inhibitory , GABAergic neurons . In addition , we further subdivide the inhibitory population in two sub-classes , I1 and I2 ( with N I 1 = 0 . 35 N I and N I 2 = 0 . 65 N I ) , corresponding to fast-spiking and non-fast-spiking interneurons , respectively ( see Neuronal properties ) . Accordingly , there are nine different synapse types ( all possible connections between neuronal populations ) , with distinct , specific response properties ( see Synaptic properties ) . Similarly , there are nine connection probabilities from which random connections are drawn ( see Structural properties ) . For each of the key features of neuronal , synaptic and structural properties , we differentiate between the homogeneous case , where all properties are identical , and the heterogeneous case , where properties are drawn from appropriately chosen distributions . In this way , we can tease apart the differential effects of the three sources of heterogeneity considered here: neuronal , synaptic and structural . For consistency , all the circuits’ structural ( and synaptic ) features are constrained primarily by the composition of layer 2/3 in the C2 barrel column in the mouse primary somatosensory cortex ( S1 ) , given the availability of direct , complete and significantly explored experimental datasets ( e . g . [92–94] ) . Throughout this study , and in order to isolate the effects of different sources of heterogeneity , we consider five different microcircuits: fully homogeneous ( Hom ) , structural ( Str ) , neuronal ( Neu ) or synaptic ( Syn ) heterogeneity in isolation and a fully heterogeneous circuit ( Het ) , accounting for the combined effects . In this section , we set out to quantify and evaluate the specific impact of the different forms of heterogeneity on the characteristics of population activity . To do so , we consider the circuit’s responses to an unspecific and stochastic external input , modelling cortical background / ongoing activity ( see Input specifications in Materials and methods ) . We determine and compare the circuit’s responsiveness by looking at the population rate transfer functions , as exemplified in Fig 4a for I2 neurons ( complete results are provided in S1 Fig ) , and summarize the results by the change in absolute gain ( ΔGain ) and offset ( ΔOffset ) introduced by each source of heterogeneity , relative to the homogeneous condition ( Fig 4b ) . All heterogeneous conditions , particularly neuronal and synaptic , cause a slight offset for all neuron types ( more significant for I2 neurons ) , making them more responsive ( firing at lower input rates ) but the effect is not substantial ( Fig 4b , top ) . In most of the conditions analysed , the E population is rather unresponsive , with less than 1% of the neurons active ( Fig 4c ) and firing at rates inferior to 1 spikes/s , regardless of the input rate . While structural and neuronal heterogeneity are incapable of circumventing this effect , synaptic heterogeneity appears to be important for the network to fire at more reasonable rates ( albeit , still very sparsely ) , resulting in a substantial modulation of the gain of the rate transfer function ( Fig 4b , bottom ) . It should be noted that the impact of structural heterogeneity alone is mitigated by the low E rates , since the structural bias exists only within excitatory synapses or between excitatory neurons and fast-spiking interneurons ( i . e . E E , I1 E , see Table 3 ) . So , if the E population rarely fires , it is difficult to ascertain the effects of structural heterogeneity , suggesting either that its relevance pertains mostly to active states , when population activity is slightly higher , or that it is negligible at this scale . The extremely sparse firing of E neurons that we observe is consistent with physiological measurements in layer 2/3 ( e . g . [92 , 94 , 111–114] ) , but it significantly limits the degree to which we can quantify the effects of heterogeneity on population activity . So , in order to obtain a greater insight , we look at the sub-threshold responses and characteristics of membrane potential dynamics ( Fig 4d and 4e ) . Excitatory neurons are always significantly hyperpolarized , with their mean membrane potentials kept far from threshold ( Fig 4d , blue ) and thus require much stronger depolarizing inputs to fire , compared with both inhibitory types . The inhibitory populations are , on average , much more depolarized and their membrane potentials fluctuate closer to their firing thresholds , particularly I1 ( Fig 4d , red ) . Qualitatively , the ratio of average degree of depolarization among the different populations is retained across all conditions , with I1 neurons being strongly depolarized , followed by I2 and E and is consistent with experimental reports for circuits in a state of quiet wakefulness ( Fig 4d , dashed lines ) . This feature stems directly from the electrophysiological properties of the different neuronal classes and the interactions among the 3 populations ( given that it is already observed in the homogeneous circuit ) . Both synaptic and neuronal heterogeneity greatly increase the variability in the distribution of mean membrane potentials across all the neurons and cause a slight overlap between E and I2 populations , an effect that is also consistent with experimental evidence [94] . Active synapses contribute to the total membrane conductance and cause a deviation from the resting membrane time constant [115 , 116] . This shunting effect may be mild in sparsely active circuits [117] , but it provides a form of activity-dependent modulation of single neurons’ integrative properties [118] , which constrain the circuit’s responsiveness . In the absence of synaptic input , I1 neurons have faster responses , characterized by a short baseline membrane time constant ( τ0 = Cm/gleak ≈ 10 . 7 ms ) , whereas I2 and E neurons are slower ( τ0 ≈ 22 . 3 and 25 . 1 ms , respectively ) and can thus integrate their synaptic inputs over a larger time scale ( dashed lines in Fig 4e ) . This relationship between the neuronal classes ( τeff ( I1 ) < τeff ( I2 ) < τeff ( E ) ) is a consequence of the neurons’ physiological properties and is consistent with empirical evidence [59 , 118 , 119] . However , when driven by external input , the ratio is modified and I2 neurons respond slowest , i . e . τeff ( I2 ) > τeff ( I1 ) > τeff ( E ) . The presence of heterogeneous synapses is important to ameliorate the magnitude of this shunting effect ( Fig 4e ) , which is very substantial in all conditions . It should be noted that , while the sparsity of recurrent activity ( particularly that of E neurons ) , would prompt us to expect a very minor reduction in τeff , the observed results are caused by the large synaptic input provided as background . In order to induce a functional state , engaging the circuit in active processing , we introduce an additional input signal , directly encoded as a piece-wise constant somatic current ( see Input specifications in Materials and methods ) . We began by tuning the input amplitudes ( of both background input firing rate νin and external input current ρu ) independently , for each condition , in order to approximate the relative ratio of mean firing rates among the different populations ( see S2 Fig ) , i . e . we attempt to find a combination of input parameters that allows the mean firing rates to remain within realistic bounds ( νE ∈ [0 . 5 , 5] , ν I 1 ∈ [ 10 , 25 ] , ν I 2 ∈ [ 3 , 15 ] , considering the values reported in [93 , 94 , 111 , 114] ) . We consider the circuit’s responses to this input signal as an active state , as opposed to the condition explored in the previous sections , where the circuit was driven solely by background , stochastic input ( noise ) . It is worth noting , however , that the similarities between what we call quiet and active states and their biological counterparts are limited ( see Discussion ) . In the following , we show that despite these limitations , the actively engaged circuit operates in similar dynamic regimes to its biological counterpart and maintains the key statistical features that are most likely to play a significant role in modulating the circuit’s processing capacity . In this section , we assess the microcircuit’s capacity to compute complex functions of the input signal , as described in the Materials and Methods . Note that we purposefully removed any predetermined structure in the input signal , such that the measurements reflect the properties of the system and not the acquisition of structural information in the input . If we were to consider naturalistic sensory input as the driving signal , this would not be the case . Furthermore , we intentionally focus on generic information processing as the ability the perform arbitrary transformations on an input signal and not on specific functions which might be performed by specific microcircuits .
Despite providing a significant step towards biological verisimilitude , our results demonstrate important limitations that ought to be addressed in future work . At the neuron level , and even though we consider three different neuronal populations , including two separate inhibitory classes , further sub-divisions have been reported in neocortical layer 2/3 populations , both for glutametergic [46] and , in particular , for GABAergic neurons [113 , 144 , 156–158] . It is possible that these reflect regional specializations particularly prominent in specific cortical areas ( such as the prefrontal cortical regions; [34 , 46 , 159] ) or that they represent separate instances of broader classes and can , for simplicity , be grouped together . Parameterized correctly , our choice of neuron model proved to be sufficient for the purposes of this study and allowed us to account for the most important physiological characteristics of the different neuronal classes and their relations ( see Neuronal properties ) . Such simplifying assumptions , however , are bound to miss relevant structural and functional features , particularly when it comes to specialization of inhibitory neurons and synapses [160–164] , the effects of dendritic nonlinearities and active dendritic processes [145–148] , intrinsic adaptation processes [165] , to name a few . It is also important , in future work along this direction , to consider the intricate relations between model parameters , i . e . explicitly include not only the empirical variability but also the covariance across multiple parameters ( as e . g . [101] ) . In this context , it should be pointed that the neuronal heterogeneity condition entailed a modification of a larger number of parameters than the other forms of heterogeneity , and so further work is needed to disentangle their contributions and obtain a single-parameter level comparison of their effects . Overall , our results lead us to conclude that it is important to understand the role of multiple interacting populations ( e . g . [166] ) , particularly including inhibitory sub-types and their different physiological properties and interactions , given their clearly distinct contributions . When it comes to synaptic transmission , we have focused on the specificities of instantaneous response kinetics and its inherent diversity , disregarding any form of synaptic plasticity . However , in our model , synaptic heterogeneity was shown to severely constrain the microcircuit’s processing capacity and memory ( Fig 8 ) , counteracting the benefits introduced by neuronal and structural heterogeneity . Additionally , the fact that the total measured capacity is very modest even in the best-performing systems ( only about 1% of the theoretical maximum ) , and considering the computational requirements posed on these systems in ecological conditions , this suggests it is reasonable to assume that there are important aspects of synaptic transmission that we have failed to consider , but contribute significantly to the circuit’s processing capacity . Adaptation and plasticity are likely to be important missing components , due to their critical roles in learning and memory processes [55] . Furthermore , variability in synaptic parameters , being the result of adaptive processes , is bound to reflect the circuit’s functional architecture , as demonstrated in e . g . [167] . Failure to consider the specificities of cortical connectivity is partially responsible for the absence of a substantial functional impact of synaptic heterogeneity in this study . Throughout this study , we have investigated the behaviour of our microcircuit model in two dynamic regimes , which we associated with the biological quiet and active states . However , the stimulation applied to bring the circuits into the active state was not biologically realistic , as we purposefully removed any spatiotemporal structure in order to measure the computational properties of the system and not the acquisition of structural information present in the input signal . Thus , the degree to which we are able to account for and explicitly compare empirical observations with the model is restricted and only qualitative . Moreover , whilst measuring the capacity of the network , we significantly under-sampled the space , as the results clearly demonstrate ( Fig 8a ) . A more complete set of basis functions would lead the capacity along both axes to decay to 0: as the complexity and memory requirements increase , the capacity to compute these functions decreases to negligible values in all systems . While accounting for delays of up to kmax = 100 allowed us to capture this effect ( since the memory range in all conditions is inferior to that ) , we failed to account for a sufficiently large dmax . The primary reason for this was computational cost , as our current implementation is extremely time-consuming ( see S2 Appendix ) . As a consequence , the capacity space is sub-normalized , incomplete and underestimated , due to the relatively small number of basis functions tested . Additionally , the limited sample size ( T = 105 ) may bias the individual results . While we have explored information processing capacity in a generic sense , future work along these lines would benefit from being more directed towards specific microcircuits engaged in specific computations . For example , the specific role of layer 2/3 microcircuits in primary and secondary visual cortices for long-range perceptual grouping have been systematically explored [14 , 15 , 168] and constitute a fruitful avenue for future research . The state of any given cortical microcircuit , both in terms of macroscopic spiking statistics and , particularly , membrane potential dynamics can differ dramatically between behavioural states [94 , 114 , 131 , 132] given that they require different levels of active ‘engagement’ . The three neuronal classes behave in very specific ways , with specialized response features providing differential contributions to the different circuit states . These neuron-class-specific contributions play an important role in the observed dynamics , providing a potential mechanism to support state modulations [123 , 169] . Spontaneous cortical activity during states of quiet wakefulness ( a quiescent state in which the animal is awake but the circuit is not directly engaged in active processing ) , is commonly characterized by short-lasting , large amplitude depolarizations [132 , 170 , 171] that reflect the presence of strongly synchronized excitatory inputs and resemble the dynamics observed under light anaesthesia [114 , 132 , 138 , 144] . Naturally , driven by a homogeneous Poisson process , the system does not exhibit such behaviour ( see Results section on Emergent population dynamics ) , which indicates that such effects are partially inherited by the spatiotemporal structure of the background input [172] , which in turn may reflect the structure of the sensory input [173] . Additionally , or alternatively , this may be a consequence of propagating waves of excitation [171] which are likely related to spatial connectivity features that were not taken into consideration in this study ( see Limitations and future work ) . Nevertheless , our quiet state , where the circuit is driven by background noise , highlights relevant features of population activity and their relations among different neuronal classes , emerging from the effects of the different sources of heterogeneity . The most prominent feature is the extremely sparse firing of E neurons ( Fig 4c ) , which appears to stem directly from the circuit’s composition ( homogeneous condition ) and is a robust and replicable effect emerging as a direct consequence of dense , strong and fast inhibition . While structural heterogeneity has no measurable effects , synaptic heterogeneity makes the E population more responsive and places some of these neurons closer to their firing thresholds ( Fig 4d ) . Neuronal heterogeneity , on the other hand , leads to more strongly hyperpolarized E and I1 populations , compared to all other conditions . This has the positive effect of shifting the distribution of membrane potential in the I1 population to a range that overlaps with the empirical values in [94] . However , E neurons become excessively hyperpolarized and their membrane potentials are kept farther from threshold and farther from the corresponding experimental value ( to which all other conditions provide a better match ) . Despite these differences , neuronal heterogeneity is responsible for placing all three neuronal populations operating within the range of values reported in the literature ( dashed lines in Fig 4d ) . Neocortical pyramidal neurons ( particularly in layer 2/3 ) fire very sparsely and are never driven to saturation , despite a large and constant synaptic bombardment . For this to occur , excitatory and inhibitory input currents onto each neuron must be carefully balanced such that , on average , they cancel each other , allowing the net mean input to be small and the output rates moderate [95 , 174] . Co-active and balanced excitation and inhibition thus stabilizes and shapes the circuit’s activity and must be actively maintained to allow the networks to operate in stable regimes [127 , 175 , 176] . Importantly , it also plays a critical role in active processing and computation , with the most clear experimental evidence coming form the development of input selectivity in visual and auditory cortices ( see e . g . [138 , 177 , 178] and references therein ) . We demonstrate that such balance condition is an emergent property from circuits with heterogeneous neurons , without the need for changing any of the system’s parameters . This observation may also provide a complementary mechanism by which cortical circuits are able to achieve and retain this dynamic balance , despite the large , potentially disruptive , variations introduced by other sources of heterogeneity , without necessarily requiring specific compensatory mechanisms as has been recently proposed by [86] . At any given point in time , the state of the circuit reflects a nonlinear combination of the current and past inputs , mediated by complex recurrent interactions . The state of each neuron is thus a nonlinear , fading memory function of the input ( the characteristics of which are determined by the circuit’s specificities and input encoding ) and the population state a set of N basis functions that can be linearly combined to approximate arbitrary nonlinear functions with fading memory . In that sense , these circuits are endowed with universal computing power on time invariant functions [16–18 , 134] . This is where complexity and heterogeneity play a particularly prominent role , as they can greatly extend the space of computable functions by diversifying population responses and , consequently , the richness of the circuit’s high-dimensional state-space . With specific functions in mind , circuits can be “designed” to perform certain computations by explicitly solving the credit-assignment problem , i . e . determining how each neuron ought to contribute to the computation [179] in order to achieve the desired outcome . This is typically achieved by constraining the microcircuit connectivity [180 , 181] and/or by postulating and building-in specific functionality ( e . g . efficient coding; [128 , 182] ) . The great majority of these approaches , however , assumes idealized conditions and neglects the complexities of real biophysics ( but see , e . g . [183] ) , which limits their scope and generalizability . Since we were not interested in specific functions , but in universal computational properties , instead of “designing” functional microcircuits or assuming specific computations , we sought to mimic fundamental design principles of the real neocortical microcircuitry and systematically evaluate how they affect the circuit’s computational capabilities . While this exploratory approach has its limitations , we were able to partially disentangle the computational role of complexity and heterogeneity in the microcircuit’s building blocks and pinpoint potential sources of functional specialization . Globally , the functional analysis on the computational benefits of the different sources of heterogeneity revealed the same effect: neuronal diversity , on its own , significantly boosts linear and nonlinear processing capacity and memory ( see Results sections on memory and processing capacity ) and dramatically increases its dynamic range and sensitivity . Surprisingly , and even though its effects on population activity were barely noticeable , structural heterogeneity has the second largest computational effect , particularly boosting the ability to compute highly nonlinear functions ( capacity at d = 4 was much larger than any other condition , see Fig 8 ) . The functional benefits introduced by neuronal and structural heterogeneity are not reflected in the fully heterogeneous circuit , given that synaptic heterogeneity prevents this from happening . It would be expectable and desirable that the computational benefits would combine in a way that could dramatically increase the total capacity of the most realistic condition . As discussed above , these synaptic effects likely reflect important limitations in our ability to capture their real influence in the biological system . Nevertheless , some of these results are in line with recent works on the effects of heterogeneity and complexity . In particular , the impact of structural heterogeneity in both macro- and microscopic connectivity have been the subject of recent investigations and are increasingly recognized as critical sources of functional specialization , endowing a network with broad and diverse temporal tuning [139] and providing important contributions to efficient memory storage and robust recall in attractor networks [184 , 185] . Despite limitations in our study , discussed above , our results highlight the importance of developing new theories of cortical function and dynamics based on the complex interactions of multiple neuronal sub-populations , as different neuronal classes have a non-negligible differential contribution to the circuit’s dynamics . Additionally , the prominent functional role of structural and neuronal heterogeneity suggest that they are part of a critical minimum necessary to account for computation in cortical microcircuit models as their effects appear to underlie a variety of important phenomena .
In all systems analysed , the neuronal dynamics is modelled using a common , simplified adaptive leaky integrate-and-fire scheme [98] , where the total current flow across the membrane of neuron i is governed by: C m d V i d t = - g leak ( V i ( t ) - E L ) - I i , adapt ( t ) - ∑ k ∈ syn ∑ j ∈ pre I ij k ( t ) ( 1 ) The spike times of neuron i are defined as the set F ( i ) = {tf|Vi ( tf ) ≥Vthresh} . At these times , the membrane potential is reset to the constant value V ( t ) = Vreset for all times t ∈ ( tf , tf + trefr] , after which integration is resumed as above . I ( t ) is the total synaptic current generated by inputs from all pre-synaptic neurons j ∈ pre mediated by synapse type k ∈ syn . To provide greater control over neuronal excitability properties and a more realistic account of cortical neuronal dynamics , we model intrinsic adaptation processes as proposed by [99]: τ w d I i , adapt d t = - I i , adapt + a ( V i ( t ) - E L ) + b ∑ t f ∈ F ( i ) δ ( t - t f ) ( 2 ) where the parameters a and b determine the relative contribution of sub-threshold and spike-triggered adaptation processes , respectively . The synaptic current ( I ij syn ) elicited by each spike from presynaptic neuron j is determined by the conductivity ( Grec ) of the corresponding , responsive receptors ( each synapse type being composed of a pre-determined set of receptors , see below ) : I ij syn ( t , V i ) = w ij syn ∑ k ∈ rec G ij k ( t , V i ) ( V i ( t ) - E k ) ( 3 ) The amplitude of post-synaptic currents is rescaled by the dimensionless weight parameter ( w ij syn ) , specific to each connection type and whose value was chosen , such that the PSP amplitudes matched the data reported in [92 , 93] ( see Table 3 ) . The synaptic conductivity G ij k in Eq 3 models the response of receptor k to spike arrival from pre-synaptic neuron j with a total conduction delay of d ij syn: G ij rec ( t , V i ) = ∑ t f ∈ F j g ij rec ( t - t f - d ij syn , V i ) ( 4 ) The conductance transient elicited by a single pre-synaptic event on a single post-synaptic receptor is then modelled as [99 , 107–109]: g ij rec ( t , V i ) = g ¯ rec n rec ( V i ) ( [ 1 - exp ( - t τ rise rec ) ] [ r rec exp ( - t τ decay f rec ) + ( 1 - r rec ) exp ( - t τ decay s rec ) ] Θ ( t ) ) ( 5 ) where g ¯ rec is the peak conductance of the corresponding receptor , nrec ( V ) is a voltage-gating function assuming a constant value of 1 for all receptor types , except NMDA , in which case [186]: n NMDA ( V i ) = ( 1 + [ Mg 2 + ] 3 . 57 mM exp ( - 0 . 062 V i ) ) - 1 ( 6 ) This gating function is thus used to model the voltage-dependent magnesium block at NMDA receptors . For simplicity , we assume a fixed [Mg2+] = 1 mM . The remaining parameters in Eq 5 correspond to the receptors’ characteristic time constants , namely the rise , fast and slow decay times , as well as the relative balance between fast and slow decay ( rrec ) . In order to account for the differential receptor composition and expression across different neuronal classes , all these parameters are specific for each receptor , synapse and neuron type . Consider the sparse adjacency matrix Asyn , specifying the anatomical connectivity between all neurons in source population pre and target population post ( with pre , post ∈ {E , I1 , I2} ) . The indices i , j of the nonzero entries in Asyn are independently drawn from normalized , truncated exponential distributions , with probability: P pre ( j ) = k out N pre exp ( - j k out N pre ) ( 7 ) P post ( i ) = k in N post exp ( - i k in N post ) ( 8 ) for pre- and postsynaptic neuron indices , respectively . Npre/post is the total number of pre-/postsynaptic neurons and kout/in are the parameters used to define the skewness of the out-/in-degree distributions , respectively . Setting kout/in = 0 corresponds to a random , uniform connectivity , whereas values >0 generate structured in-/out-degree distributions , with a larger variance in the number of connections per neuron . To adequately quantify the relevant functional properties of the microcircuits and the impact of the different features analysed , we employ metrics that are system-agnostic , i . e . independent from the specificities of the circuit analysed and , preferably , parameter-free such that the choices of metric parameters do not influence the measured outcome and any results obtained are unbiased and objectively reflect the circuit’s properties . Of particular interest , for the purposes of this study , is the adequate quantification of the characteristics of population dynamics , under active synaptic bombardment as well as the circuit’s capacity for stimulus processing and computation , in order to establish links between the features of population dynamics , the circuit’s composition and complexity and its ability to perform complex computations . All the work presented in this manuscript was implemented using the Neural Microcircuit Simulation and Analysis Toolkit ( NMSAT ) [193] , a python package designed to provide the first steps towards complex microcircuit benchmarking , as suggested and exemplified in this study . The core simulation engine running all the numerical simulations is NEST . Due to the specificities of this project , we used a modified version of NEST 2 . 10 . 0 [194] , which includes all the models used in this manuscript ( some of which are not available in the main release ) . A complete code package is provided in the supplementary materials that implements project-specific functionality to the framework , allowing the reproduction of all the numerical experiments presented in this manuscript . Computing resources were provided by the JARA-HPC Vergabegremium on the supercomputer JURECA [195] at Forschungszentrum Jülich . All numerical simulations were performed at a resolution of 0 . 1 ms , using the GSL implementation of the adaptive fourth-order Runge-Kutta method . | Cortical microcircuits are highly inhomogeneous dynamical systems whose information processing capacity is determined by the characteristics of its heterogeneous components and their complex interactions . The high degree of variability that characterizes macroscopic population dynamics , both during ongoing , spontaneous activity and active processing states reflects the underlying complexity and heterogeneity which has the potential to dramatically constrain the space of functions that any given circuit can compute , leading to richer and more expressive information processing systems . In this study , we identify different tentative sources of heterogeneity and assess their differential and cumulative contribution to the microcircuit’s dynamics and information processing capacity . We study these properties in a generic Layer 2/3 cortical microcircuit model , built and constrained by multiple sources of experimental data , and demonstrate that heterogeneity in neuronal properties and microconnectivity structure are important sources of functional specialization , greatly improving the circuit’s processing capacity , while capturing various important features of cortical physiology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"engineering",
"and",
"technology",
"nervous",
"system",
"information",
"processing",
"electrical",
"circuits",
"population",
"dynamics",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"learning",
"and",
"memory",
"cognition",
"memory",
"population",
"biology",
"information",
"technology",
"microcircuits",
"neuronal",
"tuning",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"cellular",
"neuroscience",
"cell",
"biology",
"synapses",
"anatomy",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"electrical",
"engineering",
"cognitive",
"science",
"neurophysiology"
] | 2019 | Leveraging heterogeneity for neural computation with fading memory in layer 2/3 cortical microcircuits |
Although the decision to proceed through cell division depends largely on the metabolic status or the size of the cell , the timing of cell division is often set by internal clocks such as the circadian clock . Light is a major cue for circadian clock entrainment , and for photosynthetic organisms it is also the main source of energy supporting cell growth prior to cell division . Little is known about how light signals are integrated in the control of S phase entry . Here , we present an integrated study of light-dependent regulation of cell division in the marine green alga Ostreococcus . During early G1 , the main genes of cell division were transcribed independently of the amount of light , and the timing of S phase did not occur prior to 6 hours after dawn . In contrast S phase commitment and the translation of a G1 A-type cyclin were dependent on the amount of light in a cAMP–dependent manner . CyclinA was shown to interact with the Retinoblastoma ( Rb ) protein during S phase . Down-regulating Rb bypassed the requirement for CyclinA and cAMP without altering the timing of S phase . Overexpression of CyclinA overrode the cAMP–dependent control of S phase entry and led to early cell division . Therefore , the Rb pathway appears to integrate light signals in the control of S phase entry in Ostreococcus , though differential transcriptional and posttranscriptional regulations of a G1 A-type cyclin . Furthermore , commitment to S phase depends on a cAMP pathway , which regulates the synthesis of CyclinA . We discuss the relative involvements of the metabolic and time/clock signals in the photoperiodic control of cell division .
The cell division cycle ( CDC ) is a highly conserved and regulated process , which controls the proliferation of unicellular organisms and development and tissue renewal in multicellular organisms . In eukaryotes the main steps of CDC progression are controlled by Cyclin Dependent Kinases ( CDKs ) . From human to algae , the metabolic status regulates cell cycle progression . Cell growth can occur during different phases of CDC depending on the organism but the main decision to progress into the cell cycle is usually made in G1 and depends on environmental conditions . It is referred to as cell cycle commitment and known as START in yeast or restriction point in mammals . Commitment has been depicted as a point , beyond which the cell is irreversibly engaged in cell cycle progression and is no longer sensitive to nutrients and also in the case of photosynthetic organisms , light availability [1] , [2] . The transcriptional regulation of cell cycle progression in S phase is controlled in mammals and plants by the E2F transcription factors and these are sequestrated by the Retinoblastoma protein ( Rb ) . In budding yeast , it is controlled by the transcription factor Swi4/6-dependent cell cycle box-Binding Factor ( SBF ) which is sequestrated by Whi5 . On phosphorylation of Rb by G1 Cyclin/CDC complexes , such as CyclinD-Cdk4 , E2F transcription factors are released leading to S phase commitment . In yeast on phosphorylation of Whi5 , by Cln3/Cdc28 , SBF is released leading to S phase commitment . In plants , a CyclinD/CDKA complex has been shown to phosphorylate a Retinoblastoma related ( RBR ) protein and overexpression of CyclinD accelerates entry into S phase and mitosis of G0 cells [3] . G1 cyclin/CDK complexes are primary targets of environmental signals and cyclin levels can be regulated at the transcriptional or the post-transcriptional level by mitogenic factors such as hormones and nutrients availability [4]–[7] . In animals and yeast , Rb and Whi5 respectively are critical players in linking cell size or metabolic status to cell cycle progression [8] , [9] . The gating of CDC , which restricts cell division to well defined windows of time during the day , has been described for organisms as diverse as microalgae [10] , [11] and mammals [12] . Gating of CDC ensures that cell division occurs with a daily periodicity over a wide range of environmental conditions . The timing of cell division is relatively insensitive to changes in the environment , such as nutrients or temperature and persists under constant light with a period close to 24 hours , two features of circadian regulation . The Wee1 kinase , a key regulator of G2/M transition is transcriptionally regulated by the master clock complex CLOCK/BMAL1 in mouse regenerating liver cells , illustrating the direct control of cell cycle components by the circadian clock . In addition , striking experimental evidences showed that the circadian clock and the DNA damage pathway share common regulators from animals to fungi [13]–[15] . However , more experimental data is needed to unravel cross-talks between circadian , metabolic and cell cycle controls in the absence of injury or stress . Unicellular algae such as Chlamydomonas or Euglena are very useful organisms to dissect the light-dependent regulation of cell division in photosynthetic organisms because cell division can be synchronized by light/dark cycles . In Chlamydomonas , commitment takes place in G1 whereas in Euglena the light-dependent control of CDC operates mainly in G2 but also at G1/S and S/G2 transitions [16] . In Chlamydomonas , cell division was shown to be under circadian control [17] but also to depend on the amount of light available for photosynthesis [1] . Until recently tools for gene function analysis were available only for Chlamydomonas , a microalga that exhibits multiple-fission division type . We have recently implemented molecular tools for gene function analysis in the picoeukaryotic alga Ostreococcus tauri , which divides by simple binary fission [18] . O . tauri has a very compact genome and displays very low gene redundancy [19] . A reduced set of cell cycle genes including Cyclins and Cyclin-Dependent Kinases ( CDKs ) were identified in the fully sequenced genome [20] . They encode functional CDKs and associated regulatory proteins [21] . Cell division and the transcription of the main cell cycle regulators were shown to be under circadian control and resetting by light demonstrated that the timing of cell division is mainly locked to the time of light on [22] . Here , we have performed an integrated study of light-dependent regulation of cell division in Ostreococcus , varying available light by modulating both light duration and intensity . In all conditions , the timing of cell cycle entry did not occur prior to 6 hours after dawn . No cell cycle arrest was observed outside the G1 phase . CDKA , CyclinA and Rb had patterns of expression and interactions compatible with a putative involvement in a functional Rb pathway . Cyclic AMP was necessary and sufficient for both S phase entry and CyclinA synthesis . Down-regulation of Rb or CyclinA overexpression triggered cell cycle entry under limiting light conditions demonstrating the antagonistic roles of cAMP and Rb in a “metabolic checkpoint” . Moreover , overexpression of CyclinA advanced the timing of S phase entry . Our work illustrates how combined light intensity-dependent and time-dependent signals regulate S phase entry and give insight into the role of a G1 cyclin in the light-dependent control of cell cycle progression .
Cells entrained under 12 hours light , 12 hours dark cycles ( LD 12 , 12 ) at 35 µmol . quanta . m−2 . s−1 were in G1 phase at dawn ( Time 0 ) ( Figure 1 ) . They were submitted to various light intensities and durations from Time 0 to modulate the amount of light provided ( Figure 1A ) . Under these conditions , light is a source of energy for photosynthesis , that is required for cell growth prior to cell division ( commitment ) but it can also act as a signal ( timer or clock ) controlling the timing of cell cycle events . Estimation of DNA content by flow cytometry allowed monitoring of S phase as previously described [21] , [22] . Cells in S phase were detected between 6 hours and 14 hours after light on ( Time 6 and Time 14 ) . Depending on the light intensity and duration , the cell population underwent from 0 to more than 1 division as determined by cell counting ( Figure S1 ) . G2 and M phases are very short in Ostreococcus as estimated from naturally or artificially synchronized cell populations [21] . Therefore the number of cells in G2/M is low and difficult to estimate [21] . Furthermore for low light intensities/durations , only a few cells divided ( Figure 1 ) whereas for high light intensities/durations , two successive divisions could be observed ( Figure S1 ) making it extremely difficult to estimate the rate of cells in S and G2/M phase and to discriminate between the first and second S phases . To determine the effect of light on cell division we chose to focus on the timing of entry into the first S phase ( Figure 1 ) . At the control fluence rate ( 35 µmol . quanta . m−2 . s−1 ) , S phase was detected from 6 hours after Light on ( Time 6 ) , that is , at the same time as under the entraining LD 12 , 12 cycle . When increasing light intensity ( from 35 to 100 or 150 µmol . quanta . m−2 . s−1 ) , only 3 to 4 hours of light were required for commitment to S phase . Exposure to light for 8 hours allowed S phase progression at all tested fluence rates with a maximum of cells entering S phase for highest intensities . In all conditions , cells entering S phase , completed their cell cycle and after 24 hours the cell population was back in G1 . This suggests that the main light-dependent control of cell cycle progression occurs in G1 and that cell cycle progression is not impaired by darkness once cells are committed . Commitment to S phase was dependent both on light intensity and duration ( blue area in Figure 1B ) . For example , at 150 µmol . quanta . m−2 . s−1 the first committed cells were seen 2 to 3 hours before S phase was detected , whereas at 35 µmol . quanta . m−2 . s−1 the first cells in S phase were detected at the same time as the first committed cells ( Time 6 ) ( Figure 1B ) . For the lowest light intensity , the timing of S phase was delayed ( red area in Figure 1B ) , most likely because cells had not received enough light to commit at that time ( intersection of the blue and red area on Figure 1B ) . Together these results indicate that the timing of entry into the first S phase is gated during several hours after dawn . For the highest light intensities some cells were able to divide twice in a row ( Figure S1 ) , suggesting that the timing mechanism which gates cell division until Time 5 to Time 6 , has little effect on the timing of the second division . This is similar to the gating of division described in Chlamydomonas , which is restricted to a time window , the number of successive divisions being determined by the light conditions [1] . Limiting and non-limiting light conditions for cell division ( referred to as limiting and non-limiting conditions ) were chosen as three and eight hours respectively , of exposure to light at 100 µmol . quanta . m−2 . s−1 . At this level about 90% of an LD 12 , 12 entrained cell population divided ( see Figure S1 ) . We investigated the transcription patterns of the main cell cycle actors of cell division in limiting and non-limiting conditions ( Figure 2 ) . Transcription of Cyclin-Dependent Kinases ( CDKs ) , Cyclins and Retinoblastoma ( Rb ) were monitored by quantitative RT-PCR . CyclinB was the only transcript that was not detected in limiting conditions , whereas expression patterns of other cell cycle genes including CyclinA , CyclinD , CDKA , CDKB and Rb remained similar in both limiting and non-limiting conditions . CDKA and CyclinA transcripts were detected first , accumulating as early as two hours after light on , closely followed by Rb , CyclinD and CDKB mRNAs . Maximal transcripts levels were observed between 9 to 10 hours after light on when most of the cells were progressing through the cell cycle . In the non-limiting light condition the transcription of CyclinB started after 6 hours , when S phase had begun , suggesting that its transcription might be dependent on cell cycle progression in G1 . In contrast known and putative G1/S regulators , including CDKA , CyclinA and Rb , were not differentially expressed in limiting and non-limiting conditions , indicating that their transcriptional regulation is independent of commitment . Together Figure 1 and Figure 2 suggested that commitment occurs upon light assimilation in G1 and that it does not primarily rely on transcriptional regulations of the putative G1/S regulators identified in silico . Because , the Retinoblastoma protein ( Rb ) is well known to play a central role in the restriction point of plant and animal cells , we chose to investigate the role of Rb in G1 progression in Ostreococcus . CyclinA is the only protein exhibiting a canonical ( LXCXE ) Rb-binding site [23] and CDKA is the only CDK expressed in G1 . Thus , CDKA/CyclinA complex is the best candidate for regulating cell cycle progression in G1 . To monitor Rb , CDKA , and CyclinA protein synthesis and quantify interacting partners , we generated stable translational luciferase reporter lines Rb-Luc , CDKA-Luc and CyclinA-Luc in the pOtLuc vector [18] . Estimation of the recombinant protein synthesis was achieved through luminescence measurement from either whole protein extracts or affinity-purified proteins . The human p9CKShs1 referred to as P9 was used to specifically purify CDKA [21] . An anti-CyclinA antibody was used for immunoprecipitation of CyclinA and associated proteins . Luminescence patterns measured in extracts from CDKA-Luc and CyclinA-Luc lines were similar to that of CDKA and CyclinA profiles as determined by western blot , demonstrating that in our experiments luciferase translational fusions ( Figure 3A ) reflected the expression patterns of these proteins ( Figure 3B ) , which in the case of CyclinA resulted mainly from protein de novo synthesis since endogenous CyclinA was no detected at Time 0 ( Figure 3A and 3B ) . In non-limiting conditions , CyclinA-Luc accumulated from 4 hours after light on ( Figure 4A ) . Similar profiles of CyclinA-Luc were obtained from raw extracts or P9-purified complexes ( CDKA/CyclinA-Luc ) ( Figure 4A ) . Significantly , CyclinA-Luc protein was found to be bound to CDKA from Time 4 that is , as soon as CyclinA-Luc was detected in raw extracts . Conversely , CyclinA/CDKA-Luc complexes were immuno-precipitated with the anti-CyclinA antibody . While a steady state level of CDKA-Luc was detected in raw extract , the amount of CDKA-Luc copurified with CyclinA followed the profile of CyclinA-Luc in raw extract ( Figure 4B ) . These results suggest that CyclinA may be a limiting factor in the formation of the CyclinA/CDKA complex , before commitment . Rb-Luc level increased from 6 hours after light on , reaching a maximum at 7 to 8 hours after light on ( Figure 4C ) . In contrast , Rb-Luc bound to CyclinA in immunoprecipitation experiments peaked 3 hours before Rb-Luc in raw extract ( Figure 4C ) . Rb-Luc purified on P9 ( bound to CDKA ) had a similar profile as Rb-Luc bound to CyclinA ( Figure 4D and 4C ) . Since CDKA-Luc was detected at a steady state level in raw extracts ( Figure 4B ) , this suggests that CyclinA might be a limiting factor in CDKA/Rb interaction early after dawn . Significantly , the amount of Rb-Luc associated to CyclinA or CDKA was highest around Time 5 to Time 6 , that is , 2 hours before CDKA/CyclinA maximal interaction . This suggests that at the light/dark transition ( Time 8 ) , Rb is released from the CyclinA/CDKA complex . CyclinA transcript and CyclinA-Luc were monitored in limiting and non-limiting conditions ( Figure 5 ) . In non-limiting conditions CyclinA-Luc was detected from 4 hours after light on , i . e . two hours after CyclinA transcript ( Figure 5A ) . No CyclinA-Luc could be detected in limiting conditions though CyclinA mRNA profile remained similar to that in non-limiting conditions ( Figure 5B ) . When the light supply was modulated by changing the fluence rate instead of the duration of illumination ( 12 hours of light ) , all profiles of CyclinA mRNA were increasing from 1 hour after light on ( Figure 5C ) . In contrast , translation products , as reported by CyclinA-Luc , appeared later for lower fluence rates ( Figure 5D ) . These results indicate that the synthesis of CyclinA protein , unlike the CyclinA transcription , is regulated by the light conditions in a manner very similar to S phase commitment ( e . g . Figure 1 ) . To gain insight into the signal transduction pathway leading to the light-dependent regulation of S phase entry , we chose to investigate the involvement of cAMP known to be an important signaling component for cell cycle progression [24]–[26] . Monitoring of cAMP level in cells exposed to various light intensities after LD 12 , 12 entrainment revealed that the peak of cAMP occurred earlier and/or had higher amplitudes for high fluence rates ( Figure S2 ) . This suggests a possible correlation between cAMP level , CyclinA synthesis and S phase commitment since cells were committed sooner at high fluence rates ( e . g . Figure 1 ) . Under non-limiting conditions , cAMP increased immediately after light on and returned to a basal level before S phase was detected ( Figure 6A and 6B ) . We used a pharmacological approach to evaluate the role of cAMP in the synthesis of CyclinA and the control of S phase . Indomethacin and forskolin have been reported to inhibit and activate cAMP synthesis , respectively , including in photosynthetic organisms [27] , [28] . Inhibiting cAMP synthesis with indomethacin ( 30 µM ) impaired cell cycle entry and CyclinA-Luc protein synthesis without affecting CyclinA transcription ( Figure 6A–6D ) . Conversely , the adenylate cyclase activator forskolin ( 20 µM ) significantly increased cAMP levels under limiting light conditions ( Figure 6E ) . Remarkably , S phase cells as well as CyclinA-Luc protein synthesis were detected , though at a low levels , in the presence of forskolin ( Figure 6F and 6G ) . CyclinA-Luc accumulated as early as one hour after forskolin addition , i . e . two hours after light on but the first cells in were not detected before 6 hours after light on as in control cells . Our results indicate that cAMP is required for CyclinA synthesis and S phase entry . The delay between the peak of cAMP and commitment suggests that cAMP is an upstream signal in a signal transduction pathway leading ultimately to commitment rather than a direct regulator of commitment . To better understand the respective involvement of CyclinA protein and cAMP in commitment , CyclinA was ectopically expressed under the strong and constitutive Ostreococcus High Affinity Phosphate Transporter ( HAPT ) promoter in the pOtoxLuc vector ( Figure S3 ) . Two lines , referred to as CyclinA-ox , were selected based on CyclinA expression levels . In limiting conditions CyclinA was detectable , though at low levels , as early as 1 hour after light on in CyclinA-ox lines ( Figure 7A ) . In CyclinA-ox line S phase was detected under limiting light conditions as early as 3 hours after light on ( Figure 7B ) . Under non-limiting conditions S phase entry was advanced by two hours in CyclinA-ox lines compared to WT cells ( Figure 7C ) . Moreover , unlike control cells , CyclinA–ox cells treated with indomethacin were able to enter S phase ( Figure 7D ) . Under non-limiting conditions , CyclinA displayed higher levels than control cells at all time points ( Figure 7E ) . Noteworthy , the levels of CyclinA increased after dawn in CyclinA-ox lines as in control cells , suggesting that post- translational regulations operate also in overexpression lines . Higher levels of CyclinA were also detected in indomethacin-treated cells ( Figure 7E ) . Overexpression of CyclinA , therefore , bypasses the requirement for light and cAMP , allowing cells to commit and to enter into S phase with an earlier timing . To test the role of Rb in the light-dependent regulation of cell division , we generated Retinoblastoma- knockdown ( Rb-kd ) by expressing antisense Rb sequence in the pOtoxLuc vector . Three lines , with reduced levels of Rb mRNA compared to WT cells were selected ( Figure S4 ) . Under limiting conditions , Rb-kd cells were still able to progress into S phase ( Figure 8A ) . Inhibition of cAMP synthesis by indomethacin reduced the number of Rb-kd that entered S phase but a significant proportion progressed through S phase , while wild type cells remained in G1 . Thus , down-regulation of Rb bypasses the need for light and cAMP in Rb-kd as in CyclinA-ox lines . Entry into S phase occurred at the same time in Rb-kd cells as in WT under non-limiting light conditions ( Figure 8C ) . Therefore repression of Rb expression triggers commitment but has no obvious effect on the timing of S phase , unlike overexpression of CyclinA , which induces an earlier timing of cell division . Significantly , in limiting light conditions CyclinA was not detected in Rb-kd cells ( Figure 8C ) as in control cells ( e . g . Figure 7A ) . As cells in S phase were observed in Rb-kd cells , this suggests that CyclinA may not be essential for S phase entry when Rb is repressed in limiting light conditions .
We have previously shown that in Ostreococcus , the CDC is synchronized by day/night cycles and that a timing mechanism , namely the circadian clock , regulates cell division and the transcription of the main cell cycle regulators [22] . The present study aims to decipher the molecular mechanisms involved in the photoperiodic control of cell division . Varying the fluence rate and/or the duration of exposure to light modulates the length of the G1 phase . Once committed , cells resume division indicating that the light-dependent regulation of cell division occurs mainly in G1 ( Figure 1 ) . G1 phase is lengthened by as much as 3 hours ( from 6 to 9 hours ) for the lowest light intensity , suggesting that metabolic status might control commitment . Conversely , for high fluence rates , 3 to 4 hours are sufficient for commitment but S phase is not observed prior to 6 hours after dawn indicating that engagement of committed cells into S phase is gated in time . Such a light-dependent control of cell cycle progression has been described in Chlamydomonas . In early G1 , cell cycle progression is light-dependent but after commitment it becomes light-independent and cells wait for an additional 5 to 10 hours before entering S phase [1] . In contrast , in Euglena , the light-dependent regulation of cell division occurs in both G1 and G2 phases because on transfer to darkness cell cycle arrested at both stages [16] . The main cell cycle regulatory genes display similar patterns of transcription under limiting and non-limiting light conditions ( Figure 2 ) . From completion of mitosis ( about 2 to 3 hours after light off ) to the early morning , no transcription of the main cell cycle regulators is detected [22] . CyclinA is the first gene to be transcribed from one hour after light on closely followed by CDKA , CyclinD , CDKB and Rb . ( Figure 9 ) . Transcription of these genes occurs at fixed time intervals from light on independently of the commitment status of the cells , suggesting that transcription is mainly controlled by a dawn-dependent timing mechanism , which would be similar to the circadian control of cell division [22] . The main regulatory genes of cell division , including cyclins and CDKs have been shown to retain rhythmic expression under constant light [22] suggesting a circadian regulation of their transcription . We show here that CyclinB transcription is further dependent on commitment ( Figure 2 ) . The presence of an E2F-binding motif in the promoter of CyclinB , suggests that CyclinB transcription may depend on E2F transcription factor , once the cells have passed commitment . Besides being the first cyclin to be expressed soon after dawn , CyclinA is the only cyclin , which displays a fully conserved Rb-interaction motif in O . tauri . Furthermore CyclinA interacts with both CDKA and Rb protein in G1 ( Figure 4 ) . Since CDKA is the only canonical CDK present during G1 phase [21] , CDKA/CyclinA , is potentially the only CDK/cyclin complex in the Rb pathway in Ostreococcus . As soon as CyclinA protein is detected , it is found in association with both CDKA and Rb protein . In contrast , CDKA does not associate with Rb in the absence of CyclinA suggesting that CyclinA may be essential for the formation of this complex . Finally CyclinA remains complexed with CDKA several hours after the maximal interaction between Rb/CyclinA and Rb/CDKA , consistent with the CyclinA/CDKA complex being involved in the control of S phase entry . Surprisingly , the level of Rb remains high during S phase , which may appear to be in contradiction with an exclusive role for Rb in S phase progression . This suggests that Rb might also be involved later during cell cycle progression as previously reported in several organisms [29]–[31] . CyclinA overexpression or Rb down-regulation induce cell division in limiting light conditions indicating that the Retinoblastoma pathway plays an essential role in cell cycle progression at commitment ( Figure 9B and 9C ) . Such a commitment phenotype has been observed in animal knockout cells lacking the entire Retinoblastoma family , which cannot undergo growth arrest when starved of growth factors [32] and Chlamydomonas mat3 mutants cells were shown to divide at smaller size than wild type cells [33] . The best known function of Rb is to repress S phase transcription by sequestering the E2F transcription factor until Rb is phosphorylated by specific Cyclin/CDK complexes [34] . Our results would be consistent with CyclinA/CDKA controlling cell cycle progression in G1 by regulating Rb phosphorylation at commitment , Rb being a negative regulator of CyclinA/CDKA ( Figure 9A ) . No CyclinA was detected in Rb-kd cells entering early into S phase under limiting light conditions . This would suggest that other cell cycle regulators downstream of CyclinA that are not under the control of Rb are sufficient to promote S phase entry in the absence of CyclinA ( Figure 9B ) . Therefore a main function of CyclinA/CDKA may be to override the inhibitory effect of Rb in wild type cells ( Figure 9A ) . In non-limiting light conditions , the level of cAMP increases from light on and peaks before S phase entry ( Figure S2 ) , while cAMP level remains low under limiting light conditions ( Figure 6 ) . Mitogens such as the EGF are known to induce cell cycle re-entry in a cAMP dependent manner [35] . In S . cerevisiae , G1 progression is regulated by cAMP , which mediates the intracellular level of glucose [26] . In Ostrecoccus the inhibition of cAMP synthesis with indomethacin prevents cAMP accumulation and cell division under non-limiting conditions but has no effect on cell growth ( Figure 6 ) . Conversely forskolin , an activator of cAMP synthesis , triggers cell division , though at low rates , under limiting light conditions . The fact that cAMP is necessary for commitment suggests that our limiting light condition may correspond to a metabolic limitation by restricting the light energy available for photosynthesis ( Figure 9A ) . The transcription of CyclinA does not depend on light conditions but the synthesis of CyclinA protein occurs only under non-limiting light conditions . When CyclinA protein is detected earlier under high light , the cells commit sooner . While the rise in transcription of CyclinA is independent of the light conditions occurring at a fixed time after light on , the post-transcriptional regulation of CyclinA may ensure that CyclinA does not accumulate until optimal metabolic conditions allowing cell growth for commitment are met ( Figure 9A ) . Inhibition of cAMP synthesis prevents CyclinA synthesis under non-limiting light conditions and activation of cAMP synthesis triggers CyclinA synthesis consistent with CyclinA accumulation being regulated by cAMP levels . In agreement with this hypothesis , down-regulation of Rb or overexpression of CyclinA bypasses the need for cAMP for cell division ( Figure 9B and 9C ) . A similar mechanism of G1 progression by the metabolic status has been described in yeast . The differential translation rate of the G1 cyclin cln3 is dependent on a Ras-cAMP pathway , which reflects the metabolic status of the cell [26] . In a rich carbon source , cln3 appears earlier , interacts with cdc28 to phosphorylate Whi5 and induces an earlier progression through START , leading to a shortening of G1 phase . Our results suggest that in Ostreococcus cells the synthesis of CyclinA is regulated by a cAMP dependent mechanism when cells have accumulated enough light energy ( Figure 9A ) . However , the time lapse between the peak of cAMP and S phase observed in most of the light conditions suggests that cAMP does not directly control CyclinA synthesis but that it activates a downstream pathway that controls CyclinA accumulation . An alternative explanation is that cAMP regulates cell cycle progression independently of commitment , consistent with the fact that the temporal changes in cAMP level do not correlate exactly with the timing of the commitment . The timing of cell division in Ostreococcus has been shown previously to be regulated mainly by the dark-light transition and to occur in G1 because perturbations by light or dark pulses induced changes in timing of S phase entry , but not in the duration of the S , G2 and M phases [22] . In Figure 1 , we show that the timing of S phase is delayed when entrained cells are exposed to low light from dawn . In all conditions S phase is not observed before 6 hours after light on , even for high fluence rates suggesting that the timing of S phase is gated ( Figure 1 ) . Previous light-resetting experiments have shown that the timing of cell division is mainly locked to light on at dawn defining a time window in which cells do not divide [22] . Timing mechanisms of cell division have been shown to rely on clocks such as the circadian clock in animal cells , which regulates cell cycle progression during both G1 and G2 phase . [36]–[38] . In mice liver cells re-entering the cell cycle upon liver partial ablation , the circadian clock gates entry into mitosis by regulating the transcription of the Wee1 kinase , which inhibits the activity of Cyclin B1/CDC2 kinase in G2 [36] . In Ostreococcus cells , rhythmic patterns of cell division and transcription of the main cell cycle regulators persist under constant light , supporting a circadian regulation of cell division [22] . CyclinA transcription is not affected under a wide range of light conditions further suggesting that it is regulated by a timing mechanism rather than by metabolic control . This would be consistent with a clock controlling the transcription of the main cell cycle actors after dawn . Alternatively we cannot rule out the possibility that the light on signal , on its own , is sufficient to trigger the transcription of the main cell cycle regulators in G1 . Remarkably , overexpression of CyclinA induces earlier entry into S phase in limiting light conditions , suggesting at first sight that the regulation of CDKA/CyclinA activity by a clock may account for the timing of S phase ( Figure 7 ) . However Rb-kd cells enter S phase without any detectable CyclinA under limiting light conditions and in these cells the timing of S phase is normal ( Figure 8 and Figure 9B ) . As mentioned above it is possible that CyclinA/CDKA promotes S phase entry by counteracting the inhibitory effect of Rb ( Figure 9A ) . In the absence of CyclinA and Rb , S phase would be controlled by another timing mechanism . An as yet unknown player , such as a Cyclin/CDK complex would control S phase entry independently of CyclinA/CDKA and it would be negatively regulated by Rb in normal conditions . Our results also suggest that this player would be controlled by an independent timing mechanism since the timing of S phase is normal in Rb-kd lines which display low levels of Rb transcript and no detectable CyclinA protein under limiting light conditions ( Figure 9B ) . The earlier timing of S phase entry observed in CyclinA-ox lines under limiting light conditions could also be explained by a titration effect of an inhibitor such as a CDK inhibitor by CyclinA ( Figure 9C ) . It is also possible that CyclinA overexpression non-specifically induces cell cycle progression by replacing another Cyclin/CDK complex since cyclins can have overlapping functions . Alternatively , the overexpression of CyclinA may induce cell cycle progression downstream of the G1/S progression if CyclinA is also involved later during cell cycle progression as previously hypothesized [21] . Finally , activation of cAMP synthesis by forskolin induces an early synthesis of CyclinA though at low levels but the timing of S phase is not advanced . It is therefore possible that CyclinA promotes cell cycle progression in a dose-dependent manner and that the forskolin-treatment prevents CyclinA from reaching sufficient levels to promote early S phase entry . In summary , we propose a model of the light-dependent regulation of G1 phase progression , in which timing and metabolic signals are integrated in a sequential way ( Figure 9A ) . First , the transcription of several genes needed for G1 phase progression is activated , independently of the amount of light provided . Among these genes , CyclinA is one of the earliest to be transcribed after dawn , likely depending on a timing mechanism such as the circadian clock . CyclinA synthesis appears to be controlled by a cAMP-dependent pathway , most probably under metabolic control . CyclinA protein binds both CDKA and Rb , which might result eventually in the release of E2F transcription factor upon phosphorylation of Rb . Finally another timing mechanism , independent of commitment prevents entry into S phase before 6 hours after dawn . Whether this timer is the circadian clock and which cell cycle regulators are involved is currently unknown . The limited number of cell cycle regulators in Ostreococcus as well as the recent identification of circadian clock players [18] should allow this question to be addressed in the future . More generally , unraveling the molecular mechanisms of light-dependent regulation of growth and cell division in microalgae from the phytoplankton should lead to a better understanding of the physiology of these key organisms involved in carbon dioxide assimilation .
Ostreococcus tauri strain , 0TTH0595 isolated from the Thau lagoon [39] , was cultivated in filtered sterile seawater supplemented with Keller enrichment medium ( Sigma-Aldrich , Lyon , France ) . O . tauri strain was grown in aerated flasks ( Sarstedt ) at 20°C under 12 hours light/12 hours dark cycles as previously described [21] . Drugs were purchased from Sigma-Aldrich unless otherwise stated . For extraction , cells were harvested by centrifugation in conical bottles ( 10 , 000 g , 10 min , 4°C ) , after addition of pluronic ( 0 . 1% ) to the medium and stored at −80°C until extraction . Frozen cell-pellets were ground by shaking ( 45 seconds , 30 Hz , twice ) with 5 mm stainless steel beads using a TissueLyser ( Retsch , Haan , Germany ) after the appropriate buffer was added . Cell debris were removed by centrifugation ( 12 000 g , 10 min , 4°C ) . A 1 ml cell sample was fixed with 0 . 25% glutaraldehyde ( Sigma ) for 15 min at room temperature and then stored at 4°C for 1 day or frozen in liquid nitrogen and stored at −80°C . Flow cytometry analysis was performed on a FACScan flow cytometer ( FACScalibur; Becton-Dickinson , San Jose , CA ) . Cells were counted from the appropriate gate ( FL3-H versus SSC-H ) as described previously [39] . For analysis of the DNA content , whole fixed cells were stained with SYBR green I ( 3000X dilution of the commercial solution; Molecular Probes , Eugene , OR ) for 30 min , and 20 , 000 cells per sample were analyzed using the CellQuest software . Cell cycle analysis was performed with the Modfit software ( Verity Software House , Tophsam , ME ) as previously described [21] . The Graphical trapezoidal model and the fixed ratio of G2/G1 of 1 . 85 gave the best fits and were kept for cell cycle analysis in all analysis . Amplifications by PCR of CDKA , CyclinA , Retinoblastoma , full genes , including the promoter and coding region were achieved with the Triple Master polymerase mix ( Eppendorf ) . A sub-cloning step in the pGEMT vector ( Promega ) was performed first . The pOtLuc vector was used to fuse the gene in frame with luciferase enabling protein quantification via in vitro luciferase assay [18] . The pOtoxLuc vector was designed to facilitate the selection of overexpression/antisense transformants on the basis of luminescence levels produced from luciferase fused to the CCA1 promoter ( Figure S3 ) . POtoxLuc allows the expression of the sequences of interest in sense or antisense orientation under control of the strong High Affinity Phosphate Transporter promoter ( pHAPT ) . CyclinA coding sequence and antisense of 3′-end sequence of Retinoblastoma coding sequence ( from position 2964 to 1824 ) were cloned in pOtoxLuc . Overexpression of CyclinA was confirmed by western blot , and Knock-down of Retinoblastoma by quantitative RT-PCR . Transformation was performed as previously described [18] . Briefly , O . tauri was harvested by centrifugation ( 8000 g , 8 min , 10°C ) after pluronic addition ( 0 . 1% final concentration ) and cells were gently resuspended in 1 ml 1 M sorbitol . After one supplemental wash , cells were resuspended in 50 to 80 µL sorbitol ( 2 to 3×1010 cells per ml ) and incubated with 5 µL of linearised DNA ( 1 µg/µL ) before electroporation using a Bio-rad Gene Pulser apparatus ( field strength 6 kV/cm , resistor 600 Ω , capacitor 25 µF ) . Cells were transferred into culture medium for 24 h . Stable transformant colonies were selected in semi-solid medium at 0 . 2% w/v agarose ( low melting point agarose , Invitrogen ) in Keller Medium supplemented with G418 ( Calbiochem ) at 1 mg/ml concentration . Individual clones were transferred in liquid medium in 96-well microplate until they reached stationary phase ( 4 to 6 107 cell/ml ) . Luciferase reporter lines were selected on the basis of reproducible patterns of luminescence under LD conditions . Overexpressing/Antisense lines were first selected from the lines displaying the highest luminescence level and then analyzed either by quantitative RT-PCR or by western blot . RNA was extracted using RNeasy-Plus Mini kit ( Qiagen , Hilden , Germany ) following the manufacturer's instructions . Contaminating DNA was removed using Q1 RNAse-free DNAse ( Promega ) . Absence of DNA contamination was checked by PCR . Reverse transcription was performed using the PowerScript Reverse Transcriptase synthesis kit ( BD Bioscience , Palo Alto , CA ) . Real-time PCR was carried out on a LightCycler 1 . 5 ( Roche Diagnostic ) with LightCycler DNA Master SYBR Green I ( Roche Molecular Biochemicals ) . Primers were designed with LightCycler Probe Design2 software ( Roche Diagnostic , Mannhein , Germany ) . Primers are available in Table S1 . Results were analyzed using the comparative critical threshold ( ΔΔCT ) method . The O . tauri elongation factor 1α ( EF1α was used as internal reference . The analyses were performed in duplicate . Errors ( SD ) were usually below 1% . Proteins were extracted in CCLR buffer ( 100 mM potassium phosphate pH 7 . 8 , 1 mM EDTA , 1 mM DTT , 1% TritonX-100 , 10% glycerol ) . For affinity purification , protein extracts were diluted 5 time in CCLR with antiprotease without glycerol and further incubated with either p9CKShs1 sepharose beads ( Corellou , 2005 ) or specific anti-Ostreococcus CyclinA bound to protein A sepharose on a rotator at 4°C for one hour as previously described [40] . After western blotting , protein detection was achieved by enhanced chemiluminescence detection . Luminescence of translational luciferase fusion proteins ( 50 µl of protein extract ) was recorded on Centro LB 960 luminometer ( Berthold Technologies , Germany ) , 1 minute after injection of 80 µl of luciferase assay reagent buffer ( 20 mM Tricine pH 7 . 8 , 5 mM MgCl2 , 0 . 1 mM EDTA , 3 . 3 mM DTT , 270 µM coenzyme A , 500 µM luciferin , 500 µM ATP ) . The luciferase background was determined by running controls lacking the primary antibody ( e . g anti-CyclinA ) and substracted for each time point . We also checked using various amount of recombinant luciferase expressed under control of the High affinity phosphate promoter that luciferase is not immunoprecipitated by antibodies or bound to P9 . In these control experiments , the luciferase activity was below 0 . 1% of the initial luciferase activity in the extract before immunoprecipitation . For each sample , 1 ml of cell culture ( corresponding to five millions of cells ) , were extracted in HHBS buffer . cAMP measurements were performed with cAMP I HitHunter assay kit for cells in suspension ( DiscoveRx Corp . , CA ) according to the manufacturer's instructions . Luminescence was recorded in 96 wells microplate using a Centro LB 960 luminometer ( Berthold Technologies , Germany ) . A cAMP standard curve was established to quantify the cAMP levels which were normalized to the cell number as determined by flow cytometry . | Microalgae from phytoplankton play an essential role in the biogeochemical cycles through carbon dioxide assimilation in the oceans where they account for more than half of organic carbon production . Photosynthetic cells use light energy for cell growth , but light can also reset the circadian clock , which is involved in the timing of cell division . How light signals are integrated in the control of cell division remains largely unknown in photosynthetic cells . We have used the marine picoeukaryotic alga Ostreococcus to dissect the molecular mechanisms of light-dependent control of cell division . We found that the Retinoblastoma pathway integrates light signals which regulate the synthesis of CyclinA in response to cAMP . Alteration of CyclinA or Rb levels triggers cell division in limiting light conditions and bypasses the need for cAMP . In addition , CyclinA overexpression affects the timing of S phase entry . This first integrated study of light-dependent regulation of cell division in photosynthetic cells provides insight into the underlying molecular mechanisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/plant",
"genetics",
"and",
"gene",
"expression",
"cell",
"biology/cell",
"growth",
"and",
"division",
"marine",
"and",
"aquatic",
"sciences/genetics,",
"genomics,",
"and",
"barcoding",
"cell",
"biology/cell",
"signaling"
] | 2010 | Integration of Light Signals by the Retinoblastoma Pathway in the Control of S Phase Entry in the Picophytoplanktonic Cell Ostreococcus |
Correlated neuronal activity is a natural consequence of network connectivity and shared inputs to pairs of neurons , but the task-dependent modulation of correlations in relation to behavior also hints at a functional role . Correlations influence the gain of postsynaptic neurons , the amount of information encoded in the population activity and decoded by readout neurons , and synaptic plasticity . Further , it affects the power and spatial reach of extracellular signals like the local-field potential . A theory of correlated neuronal activity accounting for recurrent connectivity as well as fluctuating external sources is currently lacking . In particular , it is unclear how the recently found mechanism of active decorrelation by negative feedback on the population level affects the network response to externally applied correlated stimuli . Here , we present such an extension of the theory of correlations in stochastic binary networks . We show that ( 1 ) for homogeneous external input , the structure of correlations is mainly determined by the local recurrent connectivity , ( 2 ) homogeneous external inputs provide an additive , unspecific contribution to the correlations , ( 3 ) inhibitory feedback effectively decorrelates neuronal activity , even if neurons receive identical external inputs , and ( 4 ) identical synaptic input statistics to excitatory and to inhibitory cells increases intrinsically generated fluctuations and pairwise correlations . We further demonstrate how the accuracy of mean-field predictions can be improved by self-consistently including correlations . As a byproduct , we show that the cancellation of correlations between the summed inputs to pairs of neurons does not originate from the fast tracking of external input , but from the suppression of fluctuations on the population level by the local network . This suppression is a necessary constraint , but not sufficient to determine the structure of correlations; specifically , the structure observed at finite network size differs from the prediction based on perfect tracking , even though perfect tracking implies suppression of population fluctuations .
The spatio-temporal structure and magnitude of correlations in cortical neural activity have been subject of research for a variety of reasons: the experimentally observed task-dependent modulation of correlations points at a potential functional role . In the motor cortex of behaving monkeys , for example , synchronous action potentials appear at behaviorally relevant time points [1] . The degree of synchrony is modulated by task performance , and the precise timing of synchronous events follows a change of the behavioral protocol after a phase of re-learning . In primary visual cortex , saccades ( eye movements ) are followed by brief periods of synchronized neural firing [2] , [3] . Further , correlations and fluctuations depend on the attentive state of the animal [4] , with higher correlations and slow fluctuations observed during quiet wakefulness , and faster , uncorrelated fluctuations in the active state [5] . It is still unclear whether the observed modulation of correlations is in fact employed by the brain , or whether it is merely an epiphenomenon . Theoretical studies have suggested a number of interpretations and mechanisms of how correlated firing could be exploited: Correlations in afferent spike-train ensembles may provide a gating mechanism by modulating the gain of postsynaptic cells ( for a review , see [6] ) . Synchrony in afferent spikes ( or , more generally , synchrony in spike arrival ) can enhance the reliability of postsynaptic responses and , hence , may serve as a mechanism for a reliable activation and propagation of precise spatio-temporal spike patterns [7] , [8] , [9] , [10] . Further , it has been argued that synchronous firing could be employed to combine elementary representations into larger percepts [11] , [12] , [7] , [13] , [14] . While correlated firing may constitute the substrate for some en- and decoding schemes , it can be highly disadvantageous for others: The number of response patterns which can be triggered by a given afferent spike-train ensemble becomes maximal if these spike trains are uncorrelated [15] . In addition , correlations in the ensemble impair the ability of readout neurons to decode information reliably in the presence of noise ( see e . g . [16] , [15] , [17] ) . Recent studies have indeed shown that biological neural networks implement a number of mechanisms which can efficiently decorrelate neural activity , such as the nonlinearity of spike generation [18] , synaptic-transmission variability and failure [19] , [20] , short-term synaptic depression [20] , heterogeneity in network connectivity [21] and neuron properties [22] and the recurrent network dynamics [23] , [24] , [17] . To study the significance of experimentally observed task-dependent correlations , it is essential to provide adequate null hypotheses: Which level and structure of correlations is to be expected in the absence of any task-related stimulus or behavior ? Even in the simplest network models without time varying input , correlations in the neural activity emerge as a consequence of shared input [25] , [26] , [27] and recurrent connectivity [24] , [28] , [17] , [29] , [30] . Irrespective of the functional aspect , the spatio-temporal structure and magnitude of correlations between spike trains or membrane potentials carry valuable information about the properties of the underlying network generating these signals [26] , [28] , [31] , [29] , [30] and could therefore help constraining models of cortical networks . Further , the quantification of spike-train correlations is a prerequisite to understand how correlation sensitive synaptic plasticity rules , such as spike-timing dependent plasticity [32] , interact with the recurrent network dynamics [33] . Finally , knowledge of the expected level of correlations between synaptic inputs is crucial for the correct interpretation of extracellular signals like the local-field potential ( LFP ) [34] . Previous theoretical studies on correlations in local cortical networks provide analytical expressions for the magnitude [27] , [24] , [17] and the temporal shape [35] , [36] , [29] , [30] of average pairwise correlations , capture the influence of the connectivity on correlations [37] , [38] , [28] , [31] , [29] , [39] , and connect oscillatory network states emerging from delayed negative feedback [40] to the shape of correlation functions [30] . In particular we have shown recently that negative feedback loops , abundant in cortical networks , constitute an efficient decorrelation mechanism and therefore allow neurons to fire nearly independently despite substantial shared presynaptic input [17] ( see also [37] , [24] , [41] ) . We further pointed out that in networks of excitatory ( E ) and inhibitory ( I ) neurons , the correlations between neurons of different cell type ( EE , EI , II ) differ in both magnitude and temporal shape , even if excitatory and inhibitory neurons have identical properties and input statistics [17] , [30] . It remains unclear , however , how this cell-type specificity of correlations is affected by the connectivity of the network . The majority of previous theoretical studies on cortical circuits is restricted to local networks driven by external sources representing thalamo-cortical or cortico-cortical inputs ( e . g . [42] , [43] , [44] ) . Most of these studies emphasize the role of the local network connectivity ( e . g . [45] ) . Despite the fact that inputs from remote ( external ) areas constitute a substantial fraction of all excitatory inputs ( about [7] , see also [46] , [47] ) , their spatio-temporal structure is often abstracted by assuming that neurons in the local network are independently driven by external sources . A priori , this assumption can hardly be justified: neurons belonging to the local cortical network receive , at least to some extent , inputs from identical or overlapping remote areas , for example due to patchy ( clustered ) horizontal connectivity [48] , [49] . Hence , shared-input correlations are likely to play a role not only for local but also for external inputs . Coherent activation of neurons in remote presynaptic areas constitutes another source of correlated external input , in particular for sensory areas [50] , [5] , [51] , [4] . So far , it is largely unknown how correlated external input affects the dynamics of local cortical networks and alters correlations in their neural activity . In this article , we investigate how the magnitude and the cell-type specificity of correlations depend on i ) the connectivity in local cortical networks of finite size and ii ) the level of correlations in external inputs . Existing theories of correlations in cortical networks are not sufficient to address these questions as they either do not incorporate correlated external input [35] , [17] , [29] , [28] , [31] or assume infinitely large networks [24] . Lindner et al . [37] studied the responses of finite populations of spiking neurons receiving correlated external input , but described inhibitory feedback by a global compound process . Our work builds on the existing theory of correlations in stochastic binary networks [35] , a well-established model in the neuroscientific community [42] , [24] . This model has the advantage of requiring for its analytical treatment elementary mathematical methods only . We employ the same network structure used in the work by Renart et al . [24] which relates the mechanism of recurrent decorrelation to the fast tracking of external signals ( see [52] for a recent review ) . This choice enables us to reconsider the explanation of decorrelation by negative feedback [17] , originally shown for networks of leaky integrate-and-fire neurons , and to compare it to the findings of Renart et al . In fact , the motivation for the choice of the model arose from the review process of [17] , during which both the reviewers and the editors encouraged us to elucidate the relation of our work to the one of Renart et al . in a separate subsequent manuscript . The present work delivers this comparison . We show here that the results presented in [17] for the leaky integrate-and-fire model are in qualitative agreement with those in networks of binary neurons . The formal relationship between spiking models and the binary neuron model is established in [53] . In particular , for weak correlations it can be shown that both models map to the Ornstein-Uhlenbeck process with one important difference: The location of the effective white noise for spiking neurons is additive in the output , while for binary neurons the effective noise is low-pass filtered , or equivalently additive on the input side of the neuron . The remainder of the manuscript is organized as follows: In “Methods” , in recurrent random networks of excitatory and inhibitory cells driven by fluctuating input from an external population of finite size . We account for the fluctuations in the synaptic input to each cell , which effectively linearize the hard threshold of the neurons [54] , [24] . We further include the resulting finite-size correlations into the established mean-field description [42] , [54] to increase the accuracy of the theory . In “Results” , we first show in “Correlations are driven by intrinsic and external fluctuations” that correlations in recurrent networks are not only caused by the externally imposed correlated input , but also by intrinsically generated fluctuations of the local populations . We demonstrate that the external drive causes an overall shift of the correlations , but that their relative magnitude is mainly determined by the intrinsically generated fluctuations . In “Cancellation of input correlations” , we revisit the earlier reported phenomenon of the suppression of correlations between input currents to pairs of cells [24] and show that it is a direct consequence of the suppression of fluctuations on the population level [17] . In “Limit of infinite network size” we consider the strong coupling limit of the theory , where the network size goes to infinity to recover earlier results for inhomogeneous connectivity [24] and to extend these results to homogeneous connectivity . Subsequently , in “Influence of connectivity on the correlation structure” , we investigate in how far the reported structure of correlations is a generic feature of balanced networks and isolate parameters of the connectivity determining this structure . Finally , in “Discussion” , we summarize our results and their implications for the interpretation of experimental data , discuss the limitations of the theory , and provide an outlook of how the improved theory may serve as a further building block to understand processing of correlated activity .
We denote the activity of neuron as . The state of a binary neuron is either or , where indicates activity , inactivity [35] , [55] , [24] . The state of the network of such neurons is described by a binary vector . We denote the mean activity as , the ( zero time lag ) covariance of the activities of a pair of neurons is defined as , where is the deviation of neuron 's activity from expectation and the average is over time and realizations of the stochastic activity . The neuron model shows stochastic transitions ( at random points in time ) between the two states and controlled by transition probabilities , as illustrated in Figure 1 . Using asynchronous update [56] , in each infinitesimal interval each neuron in the network has the probability to be chosen for update [57] , where is the time constant of the neuronal dynamics . An equivalent implementation draws the time points of update independently for all neurons . For a particular neuron , the sequence of update points has exponentially distributed intervals with mean duration , i . e . update times form a Poisson process with rate . We employ the latter implementation in the globally time-driven [58] spiking simulator NEST [59] , and use a discrete time resolution for the intervals . The stochastic update constitutes a source of noise in the system . Given the -th neuron is selected for update , the probability to end in the up-state ( ) is determined by the gain function which possibly depends on the activity of all other neurons . The probability to end in the down state ( ) is . This model has been considered earlier [60] , [35] , [55] , and here we follow the notation introduced in the latter work . The stochastic system is completely characterized by the joint probability distribution in all binary variables . An example is the recurrent random network considered here ( Figure 2 ) . Knowing the joint probability distribution , arbitrary moments can be calculated , among them pairwise correlations . Here we are only concerned with the stationary state of the network . A stationary solution of implies that for each state a balance condition holds , so that the incoming and outgoing probability fluxes sum up to zero . The occupation probability of the state is then constant . We denote as the state , where the -th neuron is active ( ) , and where neuron is inactive ( ) . Since in each infinitesimal time interval at most one neuron can change state , for each given state there are possible transitions ( each corresponding to one of the neurons changing state ) . The sum of the probability fluxes into the state and out of the state must compensate to zero [61] , so ( 1 ) From this equation we derive expressions for the first and second moments by multiplying with and summing over all possible states , which leads toNote that the term denoted does not depend on the state of neuron . We use the notation for the state of the network excluding neuron , i . e . . Separating the terms in the sum over into those with and the two terms with and , we obtainwhere we obtained the first term by explicitly summing over state ( i . e . using and evaluating the sum ) . This first sum obviously vanishes . The remaining terms are of identical form with the roles of and interchanged . We hence only consider the first of them and obtain the other by symmetry . The first term simplifies towhere we denote as the average of a function with respect to the distribution . Taken together with the mirror term , we arrive at two conditions , one for the first ( , ) and one for the second ( ) moment ( 2 ) Considering the covariance with centralized variables , for one arrives at ( 3 ) This equation is identical to eq . 3 . 9 in [35] , to eqs . 3 . 12 and 3 . 13 in [55] , and to eqs . ( 19 ) – ( 22 ) in [24 , supplement] . Starting from ( 1 ) for the general case , a similar calculation as the one resulting in ( 2 ) for leads towhere we used , valid for binary variables . As in [24] we now assume a particular form for the gain function and for the coupling between neurons by specifyingwhere is the incoming synaptic weight from neuron to neuron , is the Heaviside function , and is the threshold of the activation function . For positive the neuron gets activated only if sufficient excitatory input is present and for negative the neuron is intrinsically active even in the absence of excitatory input . We denote by the summed synaptic input to the neuron , sometimes also called the “field” . Because , the variance of a binary variable is . We now aim to solve ( 2 ) for the case , i . e . the equation . In general , the right hand side depends on the fluctuations of all neurons projecting to neuron . An exact solution is therefore complicated . However , for sufficiently irregular activity in the network we assume the neurons to be approximately independent . Further assume that in a network of homogeneous populations ( same parameters , and same statistics of the incoming connections for all neurons , i . e . same number and strength of incoming connections from neurons in a given population ) the mean activity of an individual neuron can be represented by the population mean . The mean input to a neuron in population then is ( 4 ) We assumed in the last step identical synaptic amplitudes for a synapse from a neuron in population to a neuron in population . So the input to each neuron has the same mean . As a first approximation , if the mean activity in the network is not saturated , i . e . neither nor , mapping this activity back by the inverse gain function to the input , must be close to the threshold value , so ( 5 ) This relation may be solved for and to obtain a coarse estimate of the activity in the network [42] , [54] . In mean-field approximation we assume that the fluctuations of the fields of individual neurons around their mean are mutually independent , so that the fluctuations of are , in turn , caused by a sum of independent random variables and hence the variances add up to the variance of the field ( 6 ) As is a sum of typically thousands of synaptic inputs , it approaches a Gaussian distribution with mean and variance . In this approximation the mean activity in the network is the solution of ( 7 ) This equation needs to be self-consistently solved with by numerical or graphical methods in order to obtain the stationary activity , because and depend on themselves . We here employ the algorithm and from the MINPACK package , implemented in scipy ( version 0 . 9 . 0 ) [62] as the function . In general , the term in ( 3 ) couples moments of arbitrary order , resulting in a moment hierarchy [55] . Here we only determine an approximate solution . Since the single synaptic amplitudes are small , we linearize the effect of a single synaptic input . We apply the linearization to the two terms of the form on the right hand side of ( 3 ) . In the recurrent network , the activity of each neuron in the vector may be correlated to the activity of any other neuron . Therefore , the input sensed by neuron not only depends on directly , but also indirectly through the correlations of with any of the other neurons that project to neuron . We need to take this dependence into account in the linearization . Considering the effect of one particular input explicitly one getsThe first term already contains two factors and , so it takes into account second order moments . Performing the expansion for the next input would yield terms corresponding to correlations of higher order , which are neglected here . This amounts to the assumption that the remaining fluctuations in are independent of and , and we again approximate them by a Gaussian random variable with mean and variance , so . Here we used the smallness of the synaptic weight and replaced the difference by the derivative , which has the form of a susceptibility . Using the explicit expression for the Gaussian integral ( 7 ) , the susceptibility is exactly ( 8 ) The same expansion holds for the remaining inputs to cell . With , the equation for the pairwise correlations ( 3 ) in linear approximation takes the form ( 9 ) corresponding to eq . ( 6 . 8 ) in [35] and eqs . ( 31 ) – ( 33 ) in [24 , supplement] . Note , however , that the linearization used in [35] relies on the smoothness of the gain function due to additional local noise , whereas here and in [24 , supplement] a Heaviside gain function is used and only the existence of noise generated by the network itself justifies the linearization . If the input to each neuron is homogeneous , i . e . and for all neurons in population , a structurally similar equation connects the correlations averaged over disjoint pairs of neurons belonging to two ( possibly identical ) populations , with the population averaged variances ( 10 ) In deriving the last expression , we replaced variances of individual neurons and correlations between individual pairs by their respective population averages and counted the number of connections . This equation corresponds to eqs . ( 9 . 14 ) – ( 9 . 16 ) in [35] ( which lack , however , the external population , and note the typo in the first term in line 2 of eq . ( 9 . 16 ) , which should read ) and eqs . ( 36 ) in [24 , supplement] . Written in matrix form ( 10 ) takes the form ( 24 ) stated in the results sections of the present article , where we defined ( 11 ) The explicit solution of the system of equations in the second line of ( 24 ) is ( 12 ) The mean-field solution presented in “Mean-field solution” assumes that correlations among the neurons in the network are negligible . This assumption enters the expression ( 6 ) for the variance of the input to a neuron . Having determined the actual magnitude of the correlations in ( 24 ) , we are now able to state a more accurate approximation in which we take these correlations into account , modifying the expression for the variance of the field ( 13 ) This correction suggests an iterative scheme: Initially we solve the mean-field equation ( 7 ) assuming ( hence given by ( 6 ) ) . In each step of the iteration we then calculate the correlations by ( 24 ) , compute the mean-field solution of ( 7 ) and the susceptibility ( 8 ) , taking into account the correlations ( 13 ) determined in the previous step . These steps are iterated until the solution ( ) converges . We use this approach to determine the correlation structure in Figure 3 , where we iterated until the solution became invariant up to a residual absolute difference of . A comparison of the distribution of the total synaptic input at the end of the iteration with a Gaussian distribution with parameters and is shown in Figure 3D . In the previous sections we assumed the number of incoming connections to be the same for all neurons . Studying a random network in its original Erdös-Rényi [63] sense , the number of synaptic inputs to a neuron from population is a binomially distributed random number . As a consequence , the time-averaged activity differs among neurons . Since each neuron samples a random subset of inputs from a given population , we can assume that the realization of is independent of the realization of the time-averaged activity of the inputs from population . So these two contributions to the variability of the mean input add up . The number of incoming connections to a neuron in population follows a binomial distributionwhere is the connection probability and the size of the sending population . The mean value is as before , where we denote the expectation value with respect to the realization of the connectivity as . The variance of the in-degree is henceIn the following we adapt the results from [54] , [24] to the present notation . The contribution of the variability of the number of synapses to the variance of the mean input is . The contribution from the distribution of the mean activities can be expressed by the variance of the mean activity defined asThe independently drawn inputs hence contribute , as the variances of the terms add up . So together we have [54 , eq . 5 . 5–5 . 6]Using we obtain ( 14 ) The latter expression differs from [54 , eq . 5 . 7] only in the term that is absent in the work of van Vreeswijk and Sompolinsky , because they assumed the number of synapses to be Poisson distributed in the limit of sparse connectivity [54 , Appendix , ( A . 6 ) ] ( also note that their corresponds to our ) . The expression ( 14 ) is identical to [24 , supplement , eq . ( 25 ) ] . Since the variance of a binary signal with time-averaged activity is , the population-averaged variance is hence ( 15 ) So the sum of such ( uncorrelated ) signals contributes to the fluctuation of the input as ( 16 ) The contribution due to the variability of the number of synapses can be neglected in the limit of large networks [24] . With the time-averaged activity of a single cell with mean input and variance given by ( 7 ) the distribution of activity in the population is ( 17 ) The mean activity of the whole population is ( 18 ) because the penultimate line is a convolution of two Gaussian distributions , so the means and variances add up . The second moment of the population activity is ( 19 ) These expressions are identical to [24 , supplement , eqs . ( 26 ) , ( 27 ) ] . The system of equations ( 4 ) , ( 14 ) , ( 16 ) , ( 18 ) , and ( 19 ) can be solved self-consistently . We use the algorithm and of the MINPACK package , implemented in scipy ( version 0 . 9 . 0 ) [62] as the function . This yields the self-consistent solutions for and and hence the distribution of time averaged activity ( 17 ) can be obtained , shown in Figure 4F .
To explain the correlation structure observed in a network with external inputs ( Figure 2 ) , we extend the existing theory of pairwise correlations [35] to include the effect of externally imposed correlations . The global behavior of the network can be studied with the help of the mean-field equation ( 7 ) for the population-averaged mean activity ( 20 ) where the fluctuations of the input to a neuron in population are to good approximation Gaussian with the moments ( 21 ) To determine the average activities in the network , the mean-field equation ( 20 ) needs to be solved self-consistently , as the right-hand side depends on the mean activities through ( 21 ) , as explained in “Mean-field theory including finite-size correlations” . Here denotes the number of connections from population to , and their average synaptic amplitude . Once the mean activity in the network has been found , we can determine the structure of correlations . For simplicity we focus on the zero time lag correlation , , where is the deflection of neuron 's activity from baseline and is the variance of neuron 's activity . Starting from the master equation for the network of binary neurons , in “Methods” for completeness and consistency in notation we re-derive the self-consistent equation that connects the cross covariances averaged over pairs of neurons from population and and the variances averaged over neurons from population ( 22 ) The obtained inhomogeneous system of linear equations ( 24 ) reads [35] ( 23 ) Here measures the effective linearized coupling strength from population to population . It depends on the number of connections from population to , their average synaptic amplitude and the susceptibility of neurons in population . The susceptibility given by ( 8 ) quantifies the influence of fluctuation in the input to a neuron in population on the output . depends on the working point of the neurons in population . The autocorrelations , and are the inhomogeneity in the system of equations , so they drive the correlations , as pointed out earlier [35] . This is in line with the linear theories [17] , [30] for leaky integrate-and-fire model neurons , where cross-correlations are proportional to the auto-correlations; the system of equations ( 23 ) is identical to [35 , eqs . ( 9 . 14 ) – ( 9 . 16 ) ] . Note that this description holds for finite-sized networks . With the symmetry , ( 23 ) can be written in matrix form as ( 24 ) The explicit forms of the matrices are given in ( 11 ) . This system of linear equations can be solved by elementary methods . From the structure of the equations it follows , that the correlations between the external input and the activity in the network , and , are independent of the other correlations in the network . They are solely determined by the solution of the system of equations in the second line of ( 24 ) , driven by the fluctuations of the external drive . The correlations among the neurons within the network are given by the solution of the first system in ( 24 ) . They are hence driven by two terms , the fluctuations of the neurons within the network proportional to and and the correlations between the external population and the neurons in the network , and . The second line of ( 24 ) shows that all correlations depend on the size of the external population . Since the number of randomly drawn afferents per neuron from this population is constant , the mean number of shared inputs to a pair of neurons is . In the extreme case on the left of Figure 3 all neurons receive exactly identical input . If the recurrent connectivity would be absent , we would hence have perfectly correlated activity within the local network , the covariance between two neurons would be equal to their variance , in this particular network . Figure 3A shows that the covariance in the recurrent network is much smaller; on the order of . The reason is the recently reported mechanism of decorrelation [24] , explained by the negative feedback in inhibition-dominated networks [17] . Increasing the size of the external population decreases the amount of shared input , as shown in Figure 3C . In the limit where the external drive is replaced by a constant value ( visualized as point “” ) , the external drive does consequently not contribute to correlations in the network . Figure 3A shows that the relative position of the three curves does not change with . The overall offset , however , changes . This can be understood by inspecting the analytical result ( 24 ) : The solution of this system of linear equations is a superposition of two contributions . One is due to the externally imposed fluctuations , proportional to , the other is due to fluctuations generated within the local network , proportional to and . Varying the size of the external population only changes the external contribution , causing the variation in the offset , while the internal contribution , causing the splitting between the three curves , remains constant . In the extreme case ( ) , we still observe a similar structure . The slightly larger splitting is due to the reduced variance in the single neuron input , which consequently increases the susceptibility ( 8 ) . Figure 3D shows the probability distribution of the input to a neuron in population . The histogram is well approximated by a Gaussian . The first two moments of this Gaussian are and given by ( 21 ) , if correlations among the afferents are neglected . This approximation deviates from the result of direct simulation . Taking the correlations among the afferents into account affects the variance in the input according to ( 13 ) . The latter approximation is a better estimate of the input statistics , as shown in Figure 3D . This improved estimate can be accounted for in the solution of the mean-field equation ( 20 ) , which in turn affects the correlations via the susceptibility . Iterating this procedure until convergence , as explained in “Mean-field theory including finite-size correlations” , yields the semi-analytical results presented in Figure 3 . For strongly coupled networks in the limit of large network size , previous work [24] , [52] derived a balance equation for the correlations between pairs of neurons . The expressions for the correlations are approximate at finite network size and become exact for infinitely large networks . The authors show that the resulting structure of correlations amounts to a suppression of the correlations between the input currents to a pair of cells and that the population-averaged activity closely follows the fluctuations imposed by the external drive , known as fast tracking [42] . Here we revisit these three observations - the correlation structure , the input correlation , and fast tracking - from a different view point , providing an explanation based on the suppression of population rate fluctuations by negative feedback [17] . Figure 4A shows the population activities in a network of three populations for fixed numbers of neurons and otherwise identical parameters as in [24 , their Fig . 2] . Moreover , we distributed the number of incoming connections per neuron according to a binomial distribution as in the original publication . The deflections of the excitatory and the inhibitory population partly resemble those of the external drive to the network , but partly the fluctuations are independent . Our theoretical result for the correlation structure ( 24 ) is in line with this observation: the fluctuations in the network are not only driven by external input ( proportional to ) , but also by the fluctuations generated within the local populations ( proportional to and ) , so the tracking cannot be perfect in finite-sized networks . We now consider the fluctuations in the input averaged over all neurons belonging to a particular population , . We can decompose the input to the population into contributions from excitatory ( local and external ) and from inhibitory cells , and , respectively , where we used the short hand . As shown in Figure 4E , the contributions of excitation and inhibition cancel each other so that the total input fluctuates close to the threshold ( ) of the neurons: the network is in the balanced state [42] . Moreover , this cancellation not only holds for the mean value , but also for fast fluctuations , which are consequently reduced in the sum compared to the individual components and ( Figure 4E ) . We next show that this suppression of fluctuations directly implies a relation for the correlation between the inputs to a pair of individual neurons . There are two distinct contributions to this correlation , one due to common inputs shared by the pair of neurons ( both neurons assumed to belong to population ) ( 25 ) and one due to the correlations between afferents ( 26 ) Figure 4C shows these two contributions to be of opposite sign but approximately same magnitude , as already shown in [24 , supplement] and in [17] . Figure 3C shows a further decomposition of the input correlation into contributions due to the external sources and due to connections from within the local network . The sum of all components is much smaller than each individual component . This cancellation is equivalent to small fluctuations in the population-averaged input , because ( 27 ) where in the second step we used the general relation between the covariance among two population averaged signals and , the population-averaged variance , and the pairwise averaged covariances , which reads [17 , cf . eq . ( 1 ) ] ( 28 ) We have therefore shown that the cancellation of the contribution of shared input with the contribution due to the correlations among cells is equivalent to a suppression of the fluctuations in the population-averaged input signal to the population . This suppression of fluctuations in the population-averaged input is a consequence of the overall negative feedback in these networks [17]: a fluctuation of the population averaged input causes a response in network activity which is coupled back with a negative sign , counteracting its own cause and hence suppressing the fluctuation . Expression ( 27 ) is an algebraic identity showing that hence also correlations between the total inputs to a pair of cells must be suppressed . Qualitatively this property can be understood by inspecting the mean-field equation ( 7 ) for the population-averaged activities , where we linearized the gain function around the stationary mean-field solution to obtain ( 29 ) Here the noise term qualitatively describes the fluctuations caused by the stochastic update process and the external drive ( see [53] for the appropriate treatment of the noise ) . After transformation into the coordinate system of eigenvectors ( with eigenvalue ) of the effective connectivity matrix , each component fulfills the differential equationFor stability the eigenvalues must satisfy . In the example of the network shown in Figure 4 we have the two eigenvalues ( 30 ) which in the case of identical susceptibility for all populations can be expressed in terms of the synaptic weights ( 31 ) where in the second line we inserted the numerical values of Figure 4 . The fluctuations are hence suppressed so the contributions to the fluctuations on the input side are small . This explains why fluctuations of are small in networks stabilized by negative feedback . This argument also shows why the suppression of input-correlations does not rely on a balance between excitation and inhibition; it is as well observed in purely inhibitory networks of leaky integrate-and-fire neurons [17 , cf . text following eq . ( 21 ) therein] and of binary neurons [52 , eq . ( 30 ) ] , where the overall negative feedback suppresses population fluctuations in exactly the same manner , as the only appearing eigenvalue in this case is negative . Figure 5 shows the correlations in a purely inhibitory network without any external fluctuating drive . In this network the neurons are autonomously active due to a negative threshold , which , by the cancellation argument , was chosen to obtain a mean activity of about . Pairwise correlations in the finite-sized network follow from ( 23 ) to be negative , ( 32 ) and approach in the limit of strong coupling , as also shown in [52 , eq . 30] . The contributions to the input correlation follow from ( 25 ) and ( 26 ) as ( 33 ) so that for strong negative feedback the contribution due to correlations approaches . In this limit the two contributions cancel each other as in the inhibition-dominated network with excitation and inhibition . Note , however , that the presence of externally imposed fluctuations is not required for the mechanism of cancellation by negative feedback . The negative feedback suppresses also purely network generated fluctuations . For finite coupling we have , so the total currents are always positively correlated . An interesting special case is a network with homogeneous connectivity , as studied in “Correlations are driven by intrinsic and external fluctuations” , where and , shown in Figure 6 . In this symmetric case there is only one negative eigenvalue . The other eigenvalue is , so fluctuations are only mildly suppressed in direction . However , on the input side of the neurons , these fluctuations are not seen , since their contribution to the input field is by the vanishing eigenvalue . Another consequence of the vanishing eigenvalue is that the system can freely fluctuate along the eigendirection . Consequently the tracking of the external signal is much weaker in this case , as evidenced in Figure 6A . It is easy to see that the cancellation condition ( 27 ) does not uniquely determine the structure of correlations in an network , i . e . the structure of correlations in a finite network is not uniquely determined by . This is shown in Figure 4B , illustrating as an example the correlation structure predicted in the limit of infinite network size and perfect tracking [24 , supplement , eqs . 38–39] , which fulfills exactly , because this correlation structure can alternatively be derived starting from the condition for perfect tracking . The predicted structure does not coincide with the results obtained by direct simulation of the finite network . By construction and by virtue of ( 27 ) this correlation structure , however , still fulfills the cancellation condition on the input side , as visualized in Figure 4C . We show in “Limit of infinite network size” below that the deviations from direct simulation are due to the theory being strictly valid only in the limit of infinite network size , neglecting the contribution of fluctuations of the local populations ( , ) , as they appear in ( 24 ) . Formally this is apparent from [24 , eq . ( 2 ) ] and [24 , supplement eq . ( 40–41 ) ] , stating that the solution for correlations is equivalent to the network fluctuations predominantly caused by the external input , also reflected in the expression [24 , supplement eq . ( 38–39 ) ] . This can be demonstrated explicitly by setting and in ( 24 ) , resulting in a similar prediction for , as shown in Figure 4B ( plus symbol ) . The remaining deviation between the theories is due to the different susceptibilities used by the two approaches . The full theory ( 24 ) predicts the correct correlation structure independent of the connectivity matrix . In summary , the cancellation condition imposes a constraint on the structure of correlations but is not sufficient as a unique determinant . The distribution of the in-degree in Figure 4 is an additional source of variability compared to the case of fixed in-degree . It causes a distribution of the mean activity of the neurons in the network , as shown in Figure 4F . The shape of the distribution can be assessed analytically by self-consistently solving a system of equations for the first ( 18 ) and second moment ( 19 ) of the rate distribution [54] , as described in “Influence of inhomogeneity of in-degrees” . The resulting second moments ( by simulation ) and ( by simulation ) are small compared to the mean activity . For the prediction of the covariances shown in Figure 4B–D we employed the semi-analytical self-consistent solution to determine the variances . The difference to the approximate value is , however , small for low mean activity . To relate the finite-size correlations presented in the previous sections to earlier studies on the dominant contribution to correlations in the limit of infinitely large networks [24] , we here take the limit . For non-homogeneous connectivity , we recover the earlier result [24] in “Inhomogeneous connectivity” . In “Homogeneous connectivity” we show that the correlations converge to a different limit than what would be expected from the idea of fast tracking . Starting from ( 10 ) we follow [24 , supplement] and introduce the covariances between population-averaged activities as , which leads to ( 34 ) The general solution of the continuous Lyapunov equation stated in the last line can be obtained by projecting onto the set of left-sided eigenvectors of ( see e . g . [35] eq . 6 . 14 ) . Alternatively the system of linear equations ( 34 ) may be written explicitly as ( 35 ) The solution of the latter equation is given by ( 12 ) , so . We observe that the right hand side of the first line in ( 35 ) contains again two source terms , those corresponding to fluctuations caused by the external drive ( proportional to ) and those due to fluctuations generated within the network ( proportional to or ) . This motivates our definition of the two contributions and as ( 36 ) ( 37 ) which allows us to write the full solution of ( 35 ) as . We use the superscripts and to distinguish the driving sources of the fluctuations coming from outside the network ( driven by ) and coming from within the network ( driven by and ) . Comparing Figure 6B and Figure 4B , the structure of correlations is obviously different . In Figure 6B , the structure is , whereas in Figure 4B the relation is . The only difference between these two networks is in the coupling strengths and . In the following we derive a more complete picture of the determinants of the correlation structure . In order to identify the parameters that influence the fluctuations in these networks , it is instructive to study the mean-field equation for the population-averaged activities . Linearizing ( 20 ) for small deviations of the population-averaged activity from the fixed point , for large networks with the dominant term is proportional to the change of the mean , because the standard deviation is only proportional to . To linear order we hence have a coupled set of two differential equations ( 29 ) . The dynamics of this coupled set of linear differential equations is determined by the two eigenvalues of the effective connectivity ( 30 ) . Due to the presence of the leak term on the left hand side of ( 29 ) , the fixed point rate is stable only if the real parts of the eigenvalues are both smaller than . In the network with identical input statistics for all neurons the fluctuating input is characterized by the same mean and variance for each neuron . For homogeneous neuron parameters the susceptibility is hence the same for both populations . If further the number of synaptic afferents is the same for all populations , the eigenvalues can be expressed by those of the original connectivity matrix as ( 31 ) where we defined the two parameters and which control the location of the eigenvalues . In the left column of Figure 8 we keep , , and constant and vary , where we choose the maximum value by the condition and the minimum value by the condition that and , leading to and , both fulfilled if . Varying in the right column of Figure 8 , the bounds are given by the same condition that and , so , and the condition for the larger eigenvalue to stay below or equal , so . In order for the network to maintain similar mean activity , we choose the threshold of the neurons such that the cancellation condition is fulfilled for . The resulting average activity is close to this desired value of and agrees well to the analytical prediction ( 20 ) , as shown in Figure 8 A , B . The right-most point in both columns of Figure 8 where one eigenvalue vanishes , results in the same connectivity structure . This is the case for the connectivity with the symmetry and ( cf . Figure 6 ) , because in this case the population averaged connectivity matrix has two linearly dependent rows , hence a vanishing determinant and thus an eigenvalue . As observed in Figure 8C , D at this point the absolute magnitude of correlations is largest . This is intuitively clear as the network has a degree of freedom in the direction of the eigenvector belonging to the vanishing eigenvalue . In this direction the system effectively does not feel any negative feedback , so the evolution is as if the connectivity would be absent . Fluctuations in this direction are large and are only damped by the exponential relaxation of the neuronal dynamics , given by the left hand side of ( 29 ) . The time constant of these fluctuations is then solely determined by the time constant of the single neurons , as seen in Figure 6B . From the coefficients of the eigenvector we can further conclude that the fluctuations of the excitatory population are stronger by a factor than those of the inhibitory population , explaining why , and that both populations fluctuate in-phase , so , ( Figure 8C , D , right most point ) . Moving away from this point , panels C , D in Figure 8 both show that the magnitude of correlations decreases . Comparing the temporal structures of Figure 6B and Figure 4B shows that also the time scale of fluctuations decreases . The two structural parameters and affect the eigenvalues of the connectivity in a distinct manner . Changing merely shifts the real part of both eigenvalues , but leaves their relative distance constant , as seen in Figure 8E . For smaller values of the coupling among excitatory neurons becomes weaker , so their correlations are reduced . At the left most point in Figure 8C the coupling within the excitatory population vanishes , . Changing the parameter has a qualitatively different effect on the eigenvalues , as seen in Figure 8F . At , the two real eigenvalues merge and for smaller they turn into a conjugate complex pair . At the left-most point , so both couplings within the populations vanish . The system then only has coupling from to and vice versa . The conjugate complex eigenvalues show that the population activity of the system has oscillatory solutions . This is also called the PING ( pyramidal - inhibitory - gamma ) mechanism of oscillations in the gamma-range [64] . Panels C , D in Figure 8 show that for most connectivity structures the correlation structure is , in contrast to our previous finding [17] , where we studied only the symmetric case ( the right-most point ) , at which the correlation structure is . The comparison of the direct simulation to the theoretical prediction ( 24 ) in Figure 8C , D shows that the theory yields an accurate prediction of the correlation structure for all connectivity structures considered here .
The present work explains the observed pairwise correlations in a homogeneous random network of excitatory and inhibitory binary model neurons driven by an external population of finite size . On the methodological side the work is similar to the approach taken in the work of Renart et al . [24] , that starts from the microscopic Glauber dynamics of binary networks with dense and strong synaptic coupling and derives a set of self-consistent equations for the second moment of the fluctuations in the network . As in the earlier work [24] , we take into account the fluctuations due to the balanced synaptic noise in the linearization of the neuronal response [24] , [65] rather than relying on noise intrinsic to each neuron , as in the work by Ginzburg and Sompolinsky [35] . Although the theory by Ginzburg and Sompolinsky [35] was explicitly derived for binary networks that are densely , but weakly coupled , i . e . the number of synapses per neuron is and synaptic amplitudes scale as , identical equations result for the case of strong coupling , where the synaptic amplitudes decay slower than [24] . The reason for both weakly and strongly coupled networks to be describable by the same equations lies in the self-regulating property of binary neurons: Their susceptibility ( called in the present work ) inversely scales with the fluctuations in the input , , such that and hence correlations are independent of the synaptic amplitude [65] . A difference between the work of Ginzburg and Sompolinsky [35] and the work of Renart et al . [24] is , however , that the former authors assume all correlations to be equally small , whereas the latter show that the distribution of correlations is wider than their mean due to the variability in the connectivity , in particular the varying number of common inputs . The theory yields the dominant contribution to the mean value of this distribution scaling as in the limit of infinite network size . Although the asynchronous state of densely coupled networks has been described earlier [42] , [54] by a mean-field theory neglecting correlations , the main achievement of the work by Renart et al . [24] must be seen as demonstrating that the formal structure of the theory of correlations indeed admits a solution with low correlations of order and that such a solution is accompanied by the cancellation of correlations between the inputs to pairs of neurons . In particular can this state of small correlations be achieved although the contribution of shared afferents to the input correlations is of order in the strong coupling limit , in contrast to the work of [35] , where this contribution is of order . The authors of [24] employ an elegant scaling argument , taking the network size and hence the coupling to infinity , to obtain their results . In contrast , here we study these networks at finite size and obtain a theoretical prediction in good agreement with direct simulations in a large range of biologically relevant networks sizes . We further extend the framework of correlations in binary networks by an iterative procedure taking into account the finite-size fluctuations in the mean-field solution to determine the working point ( mean activity ) of the network . We find that the iteration converges to predictions for the covariance with higher accuracy than the previous method . Equipped with these methods we investigate a network driven by correlated input due to shared afferents supplied by an external population . The analytical expressions for the covariances averaged over pairs of neurons show that correlations have two components that linearly superimpose , one caused by intrinsic fluctuations generated within the local network and one caused by fluctuations due to the external population . The size of the external population controls the strength of the correlations in the external input . We find that this external input causes an offset of all pairwise correlations , which decreases with increasing external population size in proportion to the strength of the external correlations ( ) . The structure of correlations within the local network , i . e . the differences between correlations for pairs of neurons of different types , is mostly determined by the intrinsically generated fluctuations . These are proportional to the population-averaged variances and of the activity of the neurons in the local network . As a result , the structure of correlations is mostly independent of the external drive , and hence similar to the limiting case of an infinitely large external population or the case where the external drive is replaced by a DC signal with the same mean . For the other extreme , when the size of the external population equals the number of external afferents , , all neurons receive an exactly identical external signal . We show that the mechanism of decorrelation [24] , [17] still holds for these strongly correlated external signals . The resulting correlation within the network is much smaller than expected given the amount of common input . We proceed to re-investigate three observations in balanced random networks: fast tracking of external input signals [42] , [54] , the suppression of common input correlations , and small pairwise correlations to provide a view that is complementary to previous reports [24] , [17] , [52] . The lines of argument on these matters provided in the main text of [24] and in its mathematical supplement ( as well as in [52] ) differ . The main text starts at the observation that in large networks in the inhibition-dominated regime with an invertible connectivity matrix the activity exhibits fast-tracking [24 , eq . ( 2 ) ] . The authors then argue that hence positive correlations between excitatory and inhibitory synaptic currents are responsible for the decorrelation of network activity . The mathematical supplement , however , first derives the leading term of order for the pairwise correlations in the network in the limit of infinite network size [24 , supplement , eqs . 38 , 39] and then shows that fast tracking and the cancellation of input correlations are both consequences of this correlation structure . The relation of fast tracking to the structure of correlations is a novel finding in [24 , supplement , section 1 . 4] and not contained in the original report on fast tracking [42] , [54] . We here in addition show that the cancellation of correlations between the inputs to pairs of neurons is equivalent to a suppression of fluctuations of the population-averaged input . We further demonstrate how negative feedback suppresses these fluctuations . This argument is in line with the earlier explanation that correlations are suppressed by negative feedback on the population level [17] . Dominant negative feedback is a fundamental requirement for the network to stabilize its activity in the balanced state [42] . We further show that the cancellation of input correlations does not uniquely determine the structure of correlations; different structures of correlations lead to the same cancellation of correlations between the summed inputs . The cancellation of input correlations therefore only constitutes a constraint for the pairwise correlations in the network . This constraint is identically fulfilled if the network shows perfect tracking of external input , which is equivalent to completely vanishing input fluctuations [24] . We show that the correlation structure compatible with perfect tracking [24 , supplement , eqs . 38 , 39] is generally different from the structure in finite-sized networks , although both fulfill the constraint imposed by the cancellation of input correlations . Performing the limit we distinguish two cases . ( i ) For an invertible connectivity matrix , we recover the result by [24] , that in the limit of infinite network size correlations are dominated by tracking of the external signal and intrinsically generated fluctuations can be neglected; the resulting expressions for the correlations within the network [24 , supplement , eqs . 38 , 39] are lacking the locally generated fluctuations that decay faster than for invertible connectivity . However , the intermediate result [24 , supplement , eqs . 31 , 33] is identical to [35 , eq . 6 . 8] and to ( 9 ) and contains both contributions . The convergence of the correlation structure to the limiting theory appears to be slow . For the parameters given in [24] , quantitative agreement is achieved at around neurons . For the range of network sizes up to which a random network is typically considered a good model ( neurons ) , the correlation structure is dominated by intrinsic fluctuations . ( ii ) For a singular matrix , as for example resulting from statistically identical inputs to excitatory and inhibitory neurons , the contributions of external and intrinsic fluctuations both scale as . Hence the intrinsic contribution cannot be neglected even in the limit . At finite network size the observed structure of correlations generally contains contributions from both intrinsic and external fluctuations , still present in the intermediate result [24 , supplement , eqs . 31 , 33] and in [35 , eq . 6 . 8] and ( 9 ) . In particular , the external contribution dominating in infinite networks with invertible connectivity may be negligible at finite network size . We therefore conclude that the mechanism determining the correlation structure in finite networks cannot be deduced from the limit and is not given by fast tracking of the external signal . Fast tracking is rather a consequence of negative feedback . For the common but special choice of network connectivity where the synaptic weights depend only on the type of the source but not the target neuron , i . e . and [44] , we show that the locally generated fluctuations and correlations are elevated and that the activity only loosely tracks the external input . The resulting correlation structure is . To systematically investigate the dependence of the correlation structure on the network connectivity , it proves useful to parameterize the structure of the network by two measures differentially controlling the location of the eigenvalues of the connectivity matrix . We find that for a wide parameter regime the correlations change quantitatively , but the correlation structure remains invariant . The qualitative comparison with experimental observations of [51] hence only constrains the connectivity to be within the one or the other parameter regime . The networks we study here are balanced networks in the original sense as introduced in [42] , that is to say they are inhibition-dominated and the balance of excitatory and inhibitory currents on the input side to a neuron arises as a dynamic phenomenon due to dominance of negative feedback which stabilizes the mean activity . A network with a balance of excitation and inhibition built into the connectivity of the network on the other hand would correspond in our notation to setting for both receiving populations , assuming identical sizes for the excitatory and the inhibitory population . The network activity is then no longer stabilized by negative feedback , because the mean activities and can freely co-fluctuate , and , without affecting the input to other cells: is independent of . Mathematically this amounts to a two-fold degenerate vanishing eigenvalue of the effective connectivity matrix . The resulting strong fluctuations would have to be treated with different methods than presented here and would lead to strong correlations . The current work assumes that fluctuations are sufficiently small , restricting the expressions to asynchronous and irregular network states . Technically this assumption enters in form of two approximations: First , the summed input to a cell is replaced by a Gaussian fluctuating variable , valid only if pairwise correlations are weak . Second , the effect of a single synapse on the outgoing activity of a neuron is approximated to linear order allowing us to close the hierarchy of moments , as described in [55] . Throughout this work we show in addition to the obtained approximate solutions the results of simulations of the full , non-linear system . Deviations from direct simulations are stronger at lower mean activity , when the synaptic input fluctuates in the non-linear part of the effective transfer function . The best agreement of theory and simulation is hence obtained for a mean population activity close to , where means all neurons are active . For simplicity in the major parts of this work we consider networks where neurons have a fixed in-degree . In large homogeneous random networks this is often a good approximation , because the mean number of connections is , and its standard deviation declines relative to the mean . Taking into account distributed synapse numbers and the resulting distribution of the mean activity in Figure 4 and Figure 7A shows that the results are only marginally affected for low mean activity . The impact of the activity distribution on the correlation structure is more pronounced at higher mean activity , where the second moment of the activity distribution has a notable effect on the population-averaged variance . The presented work is closely related to our previous work on the correlation structure in spiking neuronal networks [17] and indeed was triggered by the review process of the latter . In [17] , we exclusively studied the symmetric connectivity structure , where excitatory and inhibitory neurons receive the same input on average . The results are qualitatively the same as those shown in Figure 6 . A difference though is , that the external input in [17] is uncorrelated , whereas here it originates from a common finite population . The cancellation condition for input correlations , also observed in vivo [50] , holds for spiking networks as well as for the binary networks studied here . For both models , negative feedback constitutes the essential mechanism underlying the suppression of fluctuations at the population level . This can be explained by a formal relationship between the two models ( see [53] ) . Our theory presents a step towards an understanding of how correlated neuronal activity in local cortical circuits is shaped by recurrence and inputs from other cortical and thalamic areas . For example the correlation between membrane potentials of pairs of neurons in somatosensory cortex of behaving mice is dominated by low-frequency oscillations during quiet wakefulness . If the animal starts whisking , these correlations significantly decrease , even if the sensory nerve fibers are cut , suggesting an internal change of brain state [5] . Our work suggests that such a dynamic reduction of correlation could come about by modulating the effective negative feedback in the network . A possible neural implementation is the increase of tonic drive to inhibitory interneurons . This hypothesis is in line with the observed faster fluctuations in the whisking state [5] . Further work is needed to verify if such a mechanism yields a quantitative explanation of the experimental observations . The network where the number of incoming external connections per neuron equals the size of the external population , cf . Figure 3 , can be regarded as a setting where all neurons receive an identical incoming stimulus . The correlations between this signal and the responses of neurons in the local network ( Figure 3C ) are smaller than in an unconnected population without local negative feedback . This can formally be seen from ( 29 ) , because negative eigenvalues of the recurrent coupling dampen the population response of the system . This suppression of correlations between stimulus and local activity hence implies weaker responses of single neurons to the driving signal . Recent experiments have shown that only a sparse subset of around 10 percent of the neurons in S1 of behaving mice responds to a sensory stimulus evoked by the active touch of a whisker with an object [4] . The subset of responding cells is determined by those neurons in which the cell specific combination of activated excitatory and inhibitory conductances drives the membrane potential above threshold . Our work suggests that negative feedback mediated among the layer 2/3 pyramidal cells , e . g . through local interneurons , should effectively reduce their correlated firing . In a biological network the negative feedback arrives with a synaptic delay and effectively reduces the low-frequency content [17] . The response of the local activity is therefore expected to depend on the spectral properties of the stimulus . Intuitively one expects responses to better lock to the stimulus for fast and narrow transients with high-frequency content . Further work is required to investigate this issue in more detail . A large number of previous studies on the dynamics of local cortical networks focuses on the effect of the local connectivity , but ignores the spatio-temporal structure of external inputs by assuming that neurons in the local network are independently driven by external ( often Poissonian ) sources . Our study shows that the input correlations of pairs of neurons in the local network are only weakly affected by additional correlations caused by shared external afferents: Even for the extreme case where all neurons in the network receive exactly identical external input ( ) , the input correlations are small and only slightly larger than those obtained for the case where neurons receive uncorrelated external input ( ; black curve in Figure 8C ) . One may therefore conclude that the approximation of uncorrelated external input is justified . In general , this may however be a hasty conclusion . Tiny changes in synaptic-input correlations have drastic effects , for example , on the power and reach of extracellular potentials [34] . For the modeling of extracellular potentials , knowledge of the spatio-temporal structure of inputs from remote areas is crucial . The theory of correlations in presence of externally impinging signals is a required building block to study correlation-sensitive synaptic plasticity [66] in recurrent networks . Understanding the emerging structure of correlations imposed by an external signal is the first step in predicting the connectivity patterns resulting from ongoing synaptic plasticity sensitive to those correlations . | The co-occurrence of action potentials of pairs of neurons within short time intervals has been known for a long time . Such synchronous events can appear time-locked to the behavior of an animal , and also theoretical considerations argue for a functional role of synchrony . Early theoretical work tried to explain correlated activity by neurons transmitting common fluctuations due to shared inputs . This , however , overestimates correlations . Recently , the recurrent connectivity of cortical networks was shown responsible for the observed low baseline correlations . Two different explanations were given: One argues that excitatory and inhibitory population activities closely follow the external inputs to the network , so that their effects on a pair of cells mutually cancel . Another explanation relies on negative recurrent feedback to suppress fluctuations in the population activity , equivalent to small correlations . In a biological neuronal network one expects both , external inputs and recurrence , to affect correlated activity . The present work extends the theoretical framework of correlations to include both contributions and explains their qualitative differences . Moreover , the study shows that the arguments of fast tracking and recurrent feedback are not equivalent , only the latter correctly predicts the cell-type specific correlations . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"physics",
"neural",
"networks",
"computational",
"neuroscience",
"biology",
"computational",
"biology",
"neuroscience"
] | 2014 | The Correlation Structure of Local Neuronal Networks Intrinsically Results from Recurrent Dynamics |
Coxiella burnetii is an intracellular pathogen that replicates in a lysosome-derived vacuole . The molecular mechanisms used by this bacterium to create a pathogen-occupied vacuole remain largely unknown . Here , we conducted a visual screen on an arrayed library of C . burnetii NMII transposon insertion mutants to identify genes required for biogenesis of a mature Coxiella-containing vacuole ( CCV ) . Mutants defective in Dot/Icm secretion system function or the PmrAB regulatory system were incapable of intracellular replication . Several mutants with intracellular growth defects were found to have insertions in genes encoding effector proteins translocated into host cells by the Dot/Icm system . These included mutants deficient in the effector proteins Cig57 , CoxCC8 and Cbu1754 . Mutants that had transposon insertions in genes important in central metabolism or encoding tRNA modification enzymes were identified based on the appearance filamentous bacteria intracellularly . Lastly , mutants that displayed a multi-vacuolar phenotype were identified . All of these mutants had a transposon insertion in the gene encoding the effector protein Cig2 . Whereas vacuoles containing wild type C . burnetii displayed robust accumulation of the autophagosome protein LC3 , the vacuoles formed by the cig2 mutant did not contain detectible amounts of LC3 . Furthermore , interfering with host autophagy during infection by wild type C . burnetii resulted in a multi-vacuolar phenotype similar to that displayed by the cig2 mutant . Thus , a functional Cig2 protein is important for interactions between the CCV and host autophagosomes and this drives a process that enhances the fusogenic properties of this pathogen-occupied organelle .
Coxiella burnetii is a highly infectious human pathogen responsible for a global zoonotic disease called Q fever . Inhalation of contaminated aerosols by humans can lead to an acute systemic illness or a more serious chronic infection that commonly presents as endocarditis [1] , [2] . The animal reservoirs for C . burnetii include domesticated livestock , and transmission to humans from these animals can lead to outbreaks of disease , such as the Q-fever epidemic that was linked to dairy goat farms in the Netherlands [2] . Phase I strains of C . burnetii produce a lipopolysaccharide molecule that has a complex O-antigen polysaccharide chain that protects the bacteria from being killed by host serum [3] . Phase II variants of C . burnetii produce a truncated O-antigen polysaccharide and can be isolated from both infected animals and bacteria cultured ex vivo [3] , [4] . Although most strains of C . burnetii exhibit phase variation and switch between phase I and phase II spontaneously , a phase II variant of the C . burnetii Nine Mile strain RSA493 called clone 4 ( NMII ) is phase locked because it has a chromosomal deletion that eliminates several genes required for the synthesis of O-antigen polysaccharide , which makes this strain incapable of causing systemic disease in guinea pig and mouse models of infection and enhances innate immune detection [3] , [4] . Nonetheless , it has been shown that the NMII strain is indistinguishable from the isogenic phase I strain ( NMI ) in tissue culture models of infection that measure the ability of C . burnetii to replicate in human cells , which include studies in primary human monocyte-derived macrophages [5] , [6] . Importantly , NMI and NMII encode the same array of virulence determinants that have evolved for manipulating cellular functions important for intracellular replication . Intracellular replication of C . burnetii requires formation of a specialized membrane-bound compartment termed the Coxiella-containing vacuole ( CCV ) . After cells internalize C . burnetii there is host-directed transport of the CCV through the endocytic pathway , which delivers the bacteria to the low pH environment of a lysosome [7] , [8] . Intracellular C . burnetii resist the hydrolytic and bactericidal activities inside the lysosome and the acidic pH of this organelle is required to stimulate C . burnetii metabolism , which enables the bacteria to survive and replicate intracellularly [9] , [10] . Although the molecular mechanisms used by C . burnetii to transform a lysosome into a replication-permissive compartment remain unclear , there is evidence that this compartment interacts with vesicles derived from the host autophagic and secretory pathways [11]–[13] . This results in a compartment containing C . burnetii that displays the host autophagy proteins LC3 and Rab24 [12] , and late endosomal/lysosomal proteins such as LAMP1 , cathepsin D and the vacuolar type H+ ATPase [14] . It has been shown recently that the CCV accumulates host cholesterol resulting in robust localization of lipid raft proteins flotilin 1 and 2 and that this vacuole is encompassed by an F-actin cage [15] , [16] . Thus , the CCV is a unique pathogen-occupied organelle that is generated upon fusion with host lysosomes . Another unique feature of the CCV is that it has the ability to fuse promiscuously with other endosomal compartments in the cell , which consumes cellular lysosomes and results in the formation of a large lysosome-derived compartment in the infected cell [17] , [18] . Co-infection studies have shown that the ability of the CCV to fuse with other endocytic compartments will promote fusion of pre-existing phagolysosomes containing inert latex-bead particles with the CCV and will also promote the fusion of vacuoles containing other pathogenic microbes with the CCV [19] , [20] . Importantly , if a cell is independently infected with multiple C . burnetii , the ability of the bacteria to stimulate homotypic fusion of lysosome-derived compartments will lead to the formation of a single CCV in the infected cell [18] . Inhibition of bacterial protein synthesis after infection prevents bacterial manipulation of endosomal dynamics and results in contraction of the spacious CCV to create a tight-fitting membrane that surrounds bacteria residing in the CCV lumen [18] . In addition to manipulating the host membrane transport and fusion pathways to produce a mature CCV , C . burnetii also promotes host viability by actively preventing apoptosis [21] , [22] . Manipulation of membrane transport and inhibition of apoptosis are both predicted to be pathogen-directed strategies that enable C . burnetii to replicate efficiently in mammalian host cells . Deciphering the unique molecular mechanisms that C . burnetii uses to manipulate the host has become possible with the development of axenic culture conditions . Formerly classified as an obligate intracellular bacterium , it has been shown that C . burnetii replicates axenically in Acidified Cysteine Citrate Media ( ACCM ) with 5% CO2 and low oxygen conditions [23] , [24] . Subsequently , genetic approaches were developed to isolate transposon-insertion mutants and mutants with targeted gene deletions [25]–[27] , which were used to demonstrate that the Dot/Icm type IVB secretion system encoded by C . burnetii is essential for intracellular replication [27]–[29] . This secretion system is genetically and functionally related to the Dot/Icm system of the human pathogen Legionella pneumophila [30]–[32] . In L . pneumophila , the Dot/Icm system facilitates intracellular replication by translocating into the host cytosol approximately 300 different effector proteins [33] . These effectors rapidly modulate the host cell biology to direct the remodeling of the Legionella-containing vacuole ( LCV ) , which prevents fusion with lysosomes and promotes fusion of secretory vesicles to create a vacuole that supports intracellular replication [33] . The biochemical function of a small proportion of these effectors has been elucidated but correlating these functions to pathogenesis is hampered by a large degree of functional redundancy both in terms of multiple paralogs with mirroring functions [34] and dissimilar effectors targeting the same host cell pathways [35] , [36] . With few exceptions , deletion of a gene encoding a L . pneumophila effector does not typically have a measurable impact on the ability of the bacterium to replicate intracellularly . It is thought that the diversity of the natural protozoan hosts L . pneumophila encounters in nature has resulted in the selection of functionally-redundant effectors that mediate survival in specific protozoan hosts [36] . The L . pneumophila Dot/Icm system initiates effector translocation immediately upon contact with a host to prevent the LCV from engaging the host endocytic pathway [37] , [38] . By contrast , the C . burnetii Dot/Icm system does not translocate effectors until the bacteria are delivered to lysosomes and become metabolically active in an acidified vacuole [39] . Given their divergent intracellular infection strategies it is predicted that there will be minimal overlap in the function of the effectors encoded by L . pneumophila when compared to C . burnetii , which is supported by the observation that few bone fide homologs of L . pneumophila effectors are encoded in the C . burnetii genome . To date , over 100 C . burnetii Dot/Icm effectors have been identified using a range of methods [28] , [40]–[46] . The translocation of the majority of these effectors was observed using L . pneumophila as a surrogate effector delivery platform . Several C . burnetii effectors have been implicated in preventing apoptosis [43] , [47] , including the ankyrin repeat-containing protein AnkG that infers an anti-apoptotic phenotype on L . pneumophila [43] . It is predicted that C . burnetii effectors will also function to control membrane traffic , as demonstrated by the effector CvpA interacting with clathrin-coated vesicles [48] , and to manipulate other aspects of the host biology important for intracellular replication . Because C . burnetii replicates exclusively inside mammalian hosts it is predicted that there will be less functional redundancy in the cohort of C . burnetii effector proteins compared to what is observed for L . pneumophila effector proteins , and that loss of single effectors may impact CCV formation . This means that it should be possible to identify effectors important for intracellular replication , and that determining the function of these effectors will increase our understanding of CCV development . Here , we conducted a visual screen on an arrayed library of random transposon insertion mutants of C . burnetii NMII to identify genes important for formation of the mature CCV . This approach was successful and resulted in the identification of genes that are important for the intracellular lifestyle of C . burnetii . The requirement for a functional Dot/Icm system in biogenesis of the CCV was evident . Insertions that inactivated genes encoding structural components of this secretion apparatus were identified in addition to insertions in genes encoding regulatory factors that govern expression of the Dot/Icm system . Importantly , these studies have identified several effector proteins that play important and distinct roles during intracellular replication . Specifically , using this approach we reveal a genetic interaction between the effector Cig2 and the host autophagy pathway . These data indicate that Cig2 function is required for robust interactions between the CCV and host autophagosomes , and that this maintains the CCV in an autolysosomal stage of maturation .
The plasmid pKM225 encoding a Himar1 TnA7 transposase was used to introduce a transposon encoding chloramphenicol resistance and a mCHERRY fluorescent protein randomly onto the genome of the C . burnetii NMII strain RSA493 . The mutagenesis procedure was optimized to reduce isolation of siblings containing identical transposon insertions and to reduce the number of spontaneous mutants defective for Dot/Icm function ( described in Materials and Methods ) . After optimization , 3 , 840 transposon insertion mutants were isolated as single clones from 40-independent pools . These clones were then arrayed and expanded in 96-well plates containing ACCM-2 . We were successful in expanding 84 . 6% of the clones ( 3 , 237 mutants ) in liquid ACCM-2 under antibiotic selection . To determine the degree to which these clones represented independent mutants with different sites of transposon insertion , we determined the site of insertion for a total of 459 mutants using a two-stage semi-degenerate PCR amplification and sequencing protocol . This analysis confirmed that isolated clones had single transposon insertions distributed randomly across the C . burnetii genome . Additionally , this analysis revealed several mutants that had transposon insertions in genes encoding known effectors of the Dot/Icm system ( Tables S1 , S2 , S3 , S4 , S5 ) . The arrayed library of C . burnetii NMII transposon mutants was analyzed using a visual assay that assessed the ability of individual mutants to form vacuoles that support intracellular replication . Specifically , HeLa 224 cells distributed in 96-well glass-bottom plates were infected with individual mutants at an MOI of approximately 500 and then the cells were incubated for 96 h . Cells were fixed and stained with antibodies specific for Coxiella ( red ) and LAMP1 ( green ) , and the DNA was labeled with Hoechst dye ( blue ) . The ability of each mutant to form a vacuole that supported intracellular replication was assessed by direct examination using fluorescence microscopy . Importantly , the parental NMII control strain and most of the C . burnetii transposon insertion mutants formed a single spacious vacuole filled with replicating bacteria ( Figure 1A ) . Additionally , of the 459 mutants for which the transposon insertion site was determined , we found that 324 mutants ( 71% ) did not display a discernable vacuole formation defect in this visual assay , which included several mutants having insertions in genes encoding known effectors of the Dot/Icm system ( Table S5 ) . Lastly , as described in detail below , many of the mutants that displayed vacuole formation defects had insertions in genes that would be predicted to affect intracellular replication . Thus , confidence was high that this screen would identify a unique cohort of genes important for C . burnetii replication in mammalian cells . There were four distinct mutant phenotypes revealed in this visual screen ( Figure 1 ) . A severe defect characterized as no detectible intracellular replication in the visual screen was observed for 74 mutants having transposon insertions that were distributed among 21 different protein-coding regions and six different intergenic regions ( Table S1 ) . At 96 h post-infection these mutants were observed as single bacteria inside of LAMP1-positive vacuoles ( Figure 1B ) . Forty-two transposon mutants displayed a reduced ability to replicate intracellularly as determined by their presence in small vacuoles containing fewer bacteria compared to vacuoles containing the control strain ( Figure 1C , Table S2 ) . Nine transposon mutants displayed filamentous growth inside of host cells ( Figure 1E , Table S3 ) , suggesting that these bacteria were under stress or defective for cell division . Lastly , there were 10 transposon mutants isolated independently that displayed a multi-vacuolar phenotype , which was characterized by the appearance of infected host cells that contained multiple vacuoles each supporting replication of C . burnetii ( Figure 1D , Table S4 ) . Importantly , every mutant we identified that displayed this multi-vacuolar phenotype had a transposon insertion in the 2 , 430 bp region encoding the protein Cig2 ( Cbu0021 ) . Mutants with transposon insertions in the genes icmL . 2 or icmD and mutants with targeted deletions of the genes dotA or dotB were shown previously to be defective for intracellular replication [27]–[29] , which established the essential role the Dot/Icm system has in C . burnetii pathogenesis . Here , we identified 66 different intracellular growth mutants harboring a transposon insertion in dot and icm loci and three mutants that were severely attenuated for intracellular replication with insertions in this region ( Figure 2A ) . The observation that many of the intracellular growth mutants have insertions in the region encoding the Dot/Icm system , as well as the extensive coverage of this region that was obtained , provides addition evidence that the arrayed mutant library contains a random distribution of transposon insertions . Additionally , this analysis indicates that spontaneous unlinked mutations that affect Dot/Icm function did not occur at a high frequency . This was a concern because if spontaneous dot and icm mutants were encountered at a high frequency then the mutant library would not be effective at identifying effectors essential for intracellular replication , and most intracellular growth phenotypes would not be complemented in trans upon introducing plasmids encoding the wild type allele of the disrupted gene . In the pool of 450 transposon mutants that were sequenced , we identified multiple insertion mutations in the coding region located between the icmQ and icmT genes ( Figure 2A ) . No defects in CCV biogenesis were observed for these mutants , indicating that the genes in this region are not essential for Dot/Icm function . Also , we determined that the hypothetical protein Cbu1651 encoded in this region was not essential for Dot/Icm function because the mutant 16-E10 had an insertion in the cbu1651 coding region but did not have a vacuole biogenesis defect . Two mutants with the transposon inserted at the 3′ end of the cbu1651 gene displayed a vacuole biogenesis defect , however , these insertions are predicted to negatively affect expression of icmX , which is a gene essential for Dot/Icm function [49] . Previously , it was demonstrated that the C . burnetii icmD gene was required for intracellular replication [29] . Additionally , studies on L . pneumophila predict that icmC , icmN , icmT and dotD should also be important for function of the C . burnetii Dot/Icm system [50]–[53] , and a recent independent study has shown that C . burnetii mutants deficient in dotD , icmC and icmN display intracellular replication defects [54] . Although complementation studies and in-frame deletion analysis was not used to more precisely determine the specific dot and icm genes that were essential for intracellular replication , the region itself was highly represented among mutants with severe intracellular growth defects , which suggested that it should be possible to identify other important genes required for intracellular replication using this library of transposon mutants . The response regulator PmrA of L . pneumophila is important for intracellular growth of this pathogen as it controls the expression of genes encoding components of the Dot/Icm system and many effectors [55] . The C . burnetii cbu1227 gene encodes a PmrA homologue [55] , which was initially annotated as QseB [30] . The prediction is that PmrA activity is controlled by the sensor kinase PmrB encoded by an adjacent gene in the operon . At least 68 promoter regions in C . burnetii contain a consensus PmrA binding site , which included five promoter regions upstream of operons that encode most of the dot and icm genes [55] . Thus , the prediction is that expression of most C . burnetii dot and icm genes will require a functional PmrAB system . Consistent with this hypothesis , among the C . burnetii mutants identified that were defective for vacuole biogenesis , we isolated three mutants with a transposon insertion in pmrA , one mutant with a transposon insertion in pmrB , and one mutant with a transposon insertion in the regulatory region upstream of pmrA ( Figure 2B ) . To determine if the Dot/Icm system was operational in mutants defective for PmrAB function we introduced a plasmid encoding a β-lactamase reporter ( BlaM ) fused to the effector protein Cbu0077 that produces the hybrid protein BlaM-77 . The BlaM-77 protein was produced in the pmrA::Tn mutant strain and effector protein translocation was assayed during host cell infection ( Figure 2C ) . This assay used the substrate CCF4-AM , which when loaded into cells fluoresces at 535 nm ( green ) following excitation at 415 nm . However , if BlaM-77 is translocated into the host cytosol during infection , the CCF4-AM molecule will be cleaved resulting in a shift in fluorescence to 460 nm ( blue ) . No BlaM-77 translocation was detected when HeLa cells were assayed 24 h after infection with the C . burnetii pmrA::Tn mutant ( Figure 2C ) . Thus , the intracellular growth defect displayed by mutants defective in PmrAB function is likely due to a defect in Dot/Icm-dependent delivery of effector proteins important for vacuole biogenesis . Many C . burnetii NMII transposon mutants had an intracellular replication defect that resulted in a significant reduction in the size of vacuoles and the number of bacteria in each vacuole . Included in this category were three transposon insertions that were predicted to result in partial loss-of-function in the Dot/Icm system . These mutants included transposon insertions in the icmS gene encoding a chaperone protein that assists in effector translocation [56] , a mutant with an insertion located in the 3′ region of icmX that would result in the production of an IcmX protein with a small C-terminal deletion , and a mutant with an insertion in the cbu1651 gene that likely affects the expression of adjacent dot and icm genes . Reduced intracellular replication was also observed in mutants having insertions in genes encoding three different effector proteins , which were Cig57 , CoxCC8 and Cbu1754 . We isolated 10 intracellular growth mutants having a transposon insertion in cig57 and the cig57::Tn mutant called 3-H3 was analyzed in detail . When intracellular replication was measured over a seven-day period , the 3-H3 strain displayed only a 5-fold increase in genome equivalents ( GEs , Figure 3 ) . Introduction of a plasmid-encoded triple FLAG-tagged version of 3×FL-Cig57 ( pFLAG:Cig57 ) into the 3-H3 cig57::Tn mutant restored intracellular replication to levels that were equivalent to wild type C . burnetii , as determined by 237-fold increase in GEs after a 7-day infection period and the appearance by immunofluorescence microscopy of large spacious vacuoles containing replicating 3-H3 ( pFLAG:Cig57 ) bacteria ( Figure 3B & 3C ) . Importantly , there were no obvious defects observed in the maturation of vacuoles containing 3-H3 as indicated by the presence of both LAMP1 ( Figure 3 ) and cathepsin D ( Figure S1 ) on the vacuoles formed by this mutant . Thus , the effector protein Cig57 likely has a role in modulating host processes important for replication that occur after C . burnetii is delivered to a lysosome-derived organelle . We identified a strain of C . burnetii having a transposon insertion in the gene cbu1780 and a strain having an insertion in the gene cbu2072 that had both displayed a severe intracellular growth defect . Both of these genes encode hypothetical proteins , which raised the possibility they might encode novel effectors . To determine if these proteins might encode effectors both Cbu1780 and Cbu2072 were tested for Dot/Icm-dependent translocation using fusion proteins having an amino-terminal BlaM reporter . The resulting BlaM-1780 and BlaM-2072 fusion proteins were produced in C . burnetii NM II and the isogenic icmL::Tn strain . The BlaM-1780 fusion protein was translocated into the host cytosol when produced in C . burnetii with a functional Dot/Icm system , as determined by a significant increase in the 460::535 nm fluorescence ratio to 24 . 62±2 . 70 . By contrast , no translocation was detected by BlaM-2072 , 1 . 07±0 . 86 . Similarly no translocation was detected for the controls , which included BlaM-1780 and BlaM-2072 produced in the Dot/Icm-deficient mutant and BlaM alone produced in the parental C . burnetii NMII strain ( Figure S2 ) . Thus , Cbu1780 is an effector protein that has an important role during infection . A striking phenotype that resulted in the appearance of multiple CCVs in HeLa cells that were infected by C . burnetii was observed for 10 independent transposon insertion mutants ( Figure 1D ) . Because individual vacuoles containing replication-competent C . burnetii will undergo homotypic fusion inside of an infected cell this phenotype suggests that these mutants were defective in promoting homotypic fusion of the CCV . All of the mutants identified that displayed this multi-vacuolar phenotype had a transposon insertion in the gene encoding the hypothetical protein Cig2 ( Cbu0021 ) , which was recently postulated to be an effector because it could be translocated by the L . pneumophila Dot/Icm system [45] . The mutants 3-C3 and 1-D12 producing normal CCVs had a transposon insertion in the neighboring gene cbu0022 and in between cbu0022 and cbu0023 , respectively . Thus , it was unlikely that the multi-vacuole phenotype displayed by the cig2::Tn insertion mutants was due to a polar effect on expression of downstream genes . When a plasmid encoding the 3×FL-Cig2 protein ( pFLAG:Cig2 ) was introduced into the cig2::Tn mutant strain 2-E1 the resulting 2-E1 ( pFLAG:Cig2 ) strain displayed mainly single vacuoles at 72 h post-infection , which was similar to host cells infected with the wild type C . burnetii NMII strain under these same conditions ( Figure 4 ) . Despite the multi-vacuole phenotype , the cig2::Tn mutants formed vacuoles that permit bacterial replication ( Figure 4A ) . Growth curves confirmed that the cig2::Tn mutants were not defective for replication in HeLa cells ( Figure 4C ) , which suggests that cig2 might encode an effector that is required uniquely for processes important for homotypic vacuole fusion . Production of a BlaM-Cig2 protein in C . burnetii NMII revealed that Cig2 was translocated during host cell infection and that translocation of BlaM-Cig2 by C . burnetii required the Dot/Icm system ( Figure 4D ) . Thus , these data indicate that the Cig2 protein is a translocated effector required for homotypic fusion of the CCV . Previous studies have revealed a multi-vacuolar phenotype when the host gene encoding Syntaxin-17 ( STX17 ) was silenced in HeLa cells and the STX17-silenced cells were then infected with NMII [8] . The similarity between the phenotype in STX17-silenced cells and the multi-vacuole phenotype observed for the cig2::Tn mutant suggested a genetic interaction between Cig2 and STX17 . Recent data has shown that STX17 has an essential role in the host process autophagy [57] , [58] , which would suggest autophagy might be required for homotypic fusion of CCVs . LC3 is a protein that is attached to autophagosomal membranes [59] , and is important for autophagosome biogenesis and the selection of intracellular cargo that will be enveloped by autophagy . Consistent with the hypothesis that autophagy may be subverted during C . burnetii infection , it has been shown that LC3 is present on the CCV [12] . To determine if autophagy is required for CCV homotypic fusion , we used siRNA to silence the genes encoding the essential autophagy factors ATG5 and ATG12 , and vacuole biogenesis was assayed by immunofluorescence microscopy . Compared to mock-transfected cells or cells where the control protein syntaxin-18 ( STX18 ) was silenced , there was a significant increase in the percentage of C . burnetii-infected cells having two or more vacuoles per cell when the genes encoding the autophagy factors ATG5 , ATG12 , or STX17 were silenced ( Fig . 5A & B ) . Thus , a functional host autophagy system is required for Cig2-dependent homotypic fusion of the CCV . The L . pneumophila Dot/Icm effector RavZ is translocated into the host cell during infection and inhibits autophagy by directly uncoupling ATG8 proteins attached to autophagosomal membranes , which includes LC3 [60] . We generated a C . burnetii strain that produces 3×FL-RavZ to determine if autophagy is important during the initial stage of infection when the Dot/Icm system is silent or during a later stage of infection when effectors are delivered into host cells . HeLa cells infected with C . burnetii producing 3×FL-RavZ had an autophagy defect as determined by the reduction in lipidated LC3-II protein when compared to uninfected cells or cells infected with C . burnetii producing the catalytically-inactive 3×FL-RavZC258A protein ( Fig . 5D ) . Importantly , most of the cells infected with C . burnetii producing functional 3×FL-RavZ displayed the multi-vacuolar phenotype defined by the presence of two or more vacuoles containing C . burnetii , whereas cells producing the catalytically inactive 3×FL-RavZC258A protein did not ( Fig . 5C & E ) . Thus , Dot/Icm-mediated delivery of 3×FL-RavZ interfered with homotypic fusion of the CCV by blocking autophagy after bacteria had been transported to a lysosome-derived compartment in the cell , which indicates that Cig2-mediated homotypic fusion of the CCV requires membranes that display lipidated ATG8 proteins . The finding that defects in host autophagy or loss-of-function mutations in cig2 both result in a multi-vacuolar phenotype suggested that C . burnetii might subvert host autophagy by a Cig2-dependent mechanism . Consistent with this hypothesis we found that the host autophagy protein LC3 was abundant on large vacuoles containing the parental NMII strain , whereas vacuoles containing the isogenic cig2::Tn mutant had a severe defect in LC3 accumulation ( Figure 6A & B ) . LC3 accumulation at the CCV was restored when a plasmid encoding 3×FL-Cig2 was introduced into the cig2::Tn mutant . To determine if Cig2 may increase autophagy flux in infected cells the autophagy rates were assessed after infection by measuring the accumulation of lipidated LC3-II in cells . There was no appreciable difference in the amounts of lipidated LC3-II detected by immunoblot analysis when cells infected with the cig2::Tn mutant strain were compared with cells infected with the parental NMII strain of C . burnetii or compared to uninfected cells ( Figure 6C ) . Similar results were observed when autophagy flux was activated through rapamycin treatment of cells and LC3-II levels were stabilized by interfering with lysosome degradation using bafilomycin A1 ( Figure 6C ) . By contrast , a drop in LC3-II levels was measured in cells infected with C . burnetii producing 3×FL-RavZ , which results from ability of this effector to deconjugate LC3-II from membranes ( Figure 6C ) . These data indicate that infection by C . burnetii does not elevate the basal rate of autophagy and that Cig2 function does not affect autophagy flux during infection . Lastly , we asked whether the CCV created by the cig2::Tn mutant was still accessible to fluid-phase endocytic transport and whether the lumen of the vacuole remained hydrolytic . This question was addressed by pulsing infected macrophages with soluble DQ Green BSA added to the extracellular medium . Endocytic transport and cleavage of DQ Green BSA by lysosomal proteases generates fluorescent peptides that permit visualization of hydrolytic organelles by fluorescence microscopy . Similar to the organelles formed by the parental NMII strain and the cig2::Tn mutant complemented with the plasmid producing 3×FLAG-Cig2 , the vacuoles containing the Cig2-deficient mutants retained the ability to cleave DQ Green BSA as indicated by the green fluorescence localized to the CCV ( Figure 6D ) , which is consistent with the finding that these vacuoles contain the lysosomal protease Cathepsin D ( Figure S1 ) . These data indicate that compared to vacuoles formed by the parental NMII strain , the lumen of the vacuole containing the cig2::Tn mutant has a similar capacity to receive endocytic cargo and hydrolyze proteins . Thus , Cig2 function is required to promote fusion of autophagosomes with the initial acidified lysosome-derived vacuole in which C . burnetii resides .
Here we employed large-scale transposon mutagenesis to create an arrayed library of 3 , 237 C . burnetii transposon insertion mutants . The C . burnetii NMII RSA 493 genome is comprised of a chromosome that is 1 , 995 , 275 bp and a 37 , 393 bp plasmid called QpH1 [30] . The total number of mutants obtained would correlate with at least one insertion for every 628 bp of DNA assuming the transposon we used inserted randomly in the genome . Given that the average size of a C . burnetii open reading frame is 849 bp most non-essential genes should be present in the library . Consistent with these calculations , our screen identified insertions in most of the dot and icm genes predicted to be non-essential for axenic growth of C . burnetii . There were , however , also 10 independent insertions isolated in the 2 , 430 bp gene cig2 , which is higher than would be predicted given random distribution of the transposon throughout the genome . Thus , we are confident that insertions in most of the genes required for intracellular infection by C . burnetii but not for replication in axenic medium were represented in this arrayed library of mutants; however , we acknowledge that there are difficulties in reaching saturation of the genome by transposon mutagenesis that could result in a several genes required for intracellular replication not being present in this library . We found that loss-of-function mutations in the PmrAB two-component regulatory system abolished intracellular replication of C . burnetii , which is consistent with independent data reported in two recent studies [54] , [61] . Thus , regulation of the Dot/Icm system and associated effectors by the PmrAB proteins is essential for intracellular replication . Reduced intracellular replication was also observed for C . burnetii with insertions in the putative regulatory genes cbu1761 and vacB . The gene cbu1761 encodes a putative sensor histidine kinase with no apparent cognate response regulator and vacB is predicted to encode an exoribonuclease called RNase R . VacB homologs in Shigella flexneri and enteroinvasive Escherichia coli have been shown to play an important role in host virulence through post-transcriptional positive regulation of plasmid-encoded virulence genes [62] , [63] . VacB is thought to mediate this regulatory control through its capacity to process mRNA . This suggests that in addition to PmrAB being required for transcription of genes in the Dot/Icm regulon that there are other virulence-associated circuits controlled by C . burnetii regulatory proteins . Results from this screen provide initial evidence that redox metabolism is important during intracellular replication of C . burnetii . A mutant severely impaired for intracellular replication had an insertion in the gene cbu2072 ( Table S1 and Figure S2 ) . The inability to detect translocation of the BlaM-Cbu2072 fusion protein into the host cytosol during C . burnetii infection indicates Cbu2072 is unlikely to be an effector protein . Bioinformatic studies predict that the Cbu2072 protein has a molecular weight of 18 . 2 kDa and limited homology with soluble pyridine nucleotide transhydrogenases ( 30% identity over 50% of the protein ) . These enzymes provide an energy-independent means to maintain homeostasis between the two redox factors NAD ( H ) and NADP ( H ) [64] . Additional evidence that redox metabolism might be critical during intracellular replication is provided by the mutants 7-G9 , 10-B8 and 23-H5 ( Table S2 ) , which have independent transposon insertions in the nadB gene encoding L-aspartate oxidase . These C . burnetii nadB mutants had a moderate intracellular replication defect . NadB catalyzes a step in the quinolinate synthetase complex that generates quinolinic acid from aspartate [65] . Quinolinic acid acts as a precursor for the pyridine nucleotide of NAD . These processes may be specifically important for intracellular replication of C . burnetii given the high oxidative stress caused by residing in a lysosome-like organelle . Several C . burnetii mutants were identified in the visual screen because they displayed a filamentous growth phenotype . Disruption of the two-component regulatory system encoded by cbu2005 and cbu2006 , cbu0745 , mnmA , ptsP and gidA resulted in filamentous replication intracellularly . The protein Cbu0745 is predicted to be the C . burnetii homolog of ribosome-associated factor Y , and the proteins MnmA and GidA are enzymes involved in tRNA modification . Three independent mutants that displayed a filamentous growth phenotype were found to have insertions in the gidA gene , and previous studies indicate that disruption of gidA in Salmonella also results in bacteria that have defect in cell division resulting in filamentation [66] , [67] . This gidA mutant phenotype has been attributed to an altered expression of genes responsible for cell division and chromosome segregation [66] . Thus , it is likely that many of the C . burnetii mutants that demonstrate filamentation have defects in fundamental cellular processes including translation and chromosome segregation that affect cell division . Specific Dot/Icm effector proteins critical for CCV biogenesis and intracellular replication of C . burnetii were identified in this visual screen . Three other recent studies have reported C . burnetii intracellular replication defects resulting from mutations in specific Dot/Icm effectors [46] , [48] , [54] . By contrast , genetic screens to isolate intracellular replication mutants in L . pneumophila identified the Dot/Icm secretion system as being critical for intracellular replication , but did not reveal effector proteins that are essential for intracellular replication . To illustrate this point , it was shown that a Legionella strain having five large chromosomal deletions that eliminated the production of 71 different effector proteins could still replicate inside macrophages [68] , which provides further evidence that there is extensive functional redundancy built into the Legionella effector repertoire and this makes it difficult to identify effectors required for virulence by screening mutants for intracellular replication defects . Thus , the identification of effector mutants with strong intracellular growth phenotypes suggests that there is slightly less functional redundancy in the C . burnetii effector repertoire compared to Legionella . However , we identified mutants having transposon insertions in genes encoding 16 different effector proteins and were unable to detect any defects in intracellular replication or vacuole morphology for these effector mutants . Thus , it remains likely that there are functionally redundant effectors that modulate some of the host functions required for intracellular replication of C . burnetii . Additionally , it is likely that some of the effectors that are encoded by C . burnetii play important roles during infection of animals even though these effectors are not required for C . burnetii replication in host cells cultured ex vivo . Hypothetically , there could be effectors that modulate inflammation by preventing detection of C . burnetii by either innate or adaptive immune surveillance that would be predicted to fall into this category . In our initial attempts at using transposon insertion mutagenesis to identify genes important for intracellular replication we were befuddled by loss-of-function mutations presumably arising spontaneously at a high frequency in dot and icm genes , which resulted in intracellular growth defects that were not linked to the site of transposon insertion . We optimized the mutagenesis protocol to reduce the probability of phenotypic differences being the result of spontaneous unlinked mutations . By either isolating multiple independent insertions in a gene where all mutants display the same phenotype or by complementing a phenotype by introducing a wild type allele on a plasmid , we demonstrate here that there are distinct phenotypes that are linked to transposon-mediated inactivation of a specific gene . However , it remains possible that some of the mutant phenotypes reported for insertion mutants described in the Supplemental Tables could be due to unlinked mutations and further studies are needed to support this initial analysis . Our data also suggest that unlinked mutations may have complicated results in a recent study where transposon insertions in genes encoding effector proteins were found to affect intracellular replication [46] . Complementation studies were not included in this study , which made it difficult to rule out the possibility that some of the transposon insertion mutants with intracellular growth defects had unlinked mutations that affect Dot/Icm function or the function of some other gene important for infection . For example , it was reported that a cbu2052 transposon insertion mutant had an intracellular replication defect , however , we obtained two independent mutants with insertions in the cbu2052 gene and immediately upstream of cbu2052 ( Table S5 ) and both of these mutants formed CCVs that were indistinguishable from the vacuoles formed by the parental strain of C . burnetii . Thus , it is important that transposon insertion phenotypes in C . burnetii are validated using either complementation or allelic replacement approaches before important functions are assigned to effector proteins . Ten intracellular replication mutants isolated in the screen were found to have independent insertions in the cig57 gene and the intracellular replication defect displayed by a cig57::Tn mutant was complemented using a plasmid encoded cig57 allele . Cig57 is highly conserved among sequenced C . burnetii strains , however , database searches did not identify other proteins with homology to Cig57 . Thus , we can conclude with high confidence that Cig57 represents a unique effector protein that has an activity that is important for C . burnetii intracellular replication . In addition to identifying mutants defective for intracellular replication , the visual screen revealed that C . burnetii cig2 mutants display a multi-vacuole phenotype . Whereas infection of a single cell by multiple C . burnetii usually leads to formation of a single vacuole due to homotypic fusion of the individual CCVs , the vacuoles containing cig2 mutants do not display the same propensity to fuse with each other inside the host cell , which results in a single host cell having multiple CCVs that each display LAMP1 and cathepsin D localization at the limiting membrane of the proteolytic lysosome-derived organelle . The cig2 gene encodes a protein with a predicted molecular weight of 92 . 9 kDa . The Cig2 protein is encoded in the genomes of all sequenced strains of C . burnetii , however , the protein does not have homologs in other organisms and there are no conserved domains that might aid in predicting the biochemical functions of this protein . Our data demonstrate that Cig2 is translocated into host cells during infection by C . burnetii using a mechanism that requires the Dot/Icm system . Additionally , it has been shown that Cig2 produced in Legionella can be translocated into host cells by the Dot/Icm system [45] . Thus , Cig2 represents a functional Dot/Icm effector protein that modulates vacuole biogenesis . The multi-vacuolar phenotype displayed by cig2::Tn mutants was similar to the multi-vacuolar phenotype displayed after STX17 was silenced and HeLa cells were infected with the parental NMII strain [8] . Why silencing of STX17 would result in a multi-vacuole phenotype was unclear initially , however , recent studies have shown that STX17 function is critical for autophagy in mammalian cells [57] , [58] . This suggested that host autophagy was required for homotypic fusion of CCVs . Indeed , our data show that silencing host genes encoding essential component of the autophagy machinery resulted in the multi-vacuole phenotype in C . burnetii-infected cells . Additionally , when the LC3-deconjugating effector RavZ was introduced into C . burnetii , the RavZ-producing C . burnetii were able to disrupt host autophagy and this resulted in a multi-vacuole phenotype . These data provide a clear phenotypic link between the host autophagy system and Cig2 function . Similar to the unregulated fusion that occurs between pre-existing phagolysosomes and the CCV in infected cells , upregulation of autophagy in mammalian cells generates large autolysosomal organelles as autophagosomes consume lysosomes through rapid fusion [69] . Importantly , independent studies have shown that LC3 associates with the CCV during vacuole biogenesis by an active process mediated by viable C . burnetii [12] , [13] . Additionally , it has been shown that the presence of LC3 on phagosome membranes will promote rapid fusion with lysosomes by a process known as LC3-associated phagocytosis [70] . Thus , we hypothesized that the reason a C . burnetii cig2 mutant displays a multi-vacuolar phenotype is because this effector is important for autophagy subversion by C . burnetii . Finding that there is defect in the localization of LC3 to vacuoles formed by Cig2-deficient C . burnetii supports this hypothesis . Finding that the rates of autophagy were similar following infection of cells with NMII or the isogenic cig2::Tn mutant indicates that Cig2 does not stimulate a general upregulation of autophagy flux in the infected cells . This suggests that Cig2 function is required to promote fusion of autophagosomes that are generated at a basal level in the infected cells with the CCV . Based on these data , we propose a model whereby autophagy subversion by Cig2 is required to constitutively promote the fusion of autophagosomes with the CCV during infection . This would enable Atg8 proteins such as LC3 to be maintained on the CCV membrane and keep the CCV in autolysosomal stage of maturation . We postulate that by locking the CCV in an autolysosomal stage of maturation this vacuole would remain highly fusogenic , and this would promote homotypic fusion and fusion of the CCV with other lysosome-derived organelles in the cell . The result would be formation of a spacious CCV and the fusion of lysosome-derived organelles containing other bacteria or inert particles with the CCV . Determining whether this model is correct will require elucidating the biochemical function of Cig2 and a better understanding of the role autophagy subversion plays in generating the vacuole that C . burnetii occupies .
Plaque purified C . burnetii Nine Mile phase II ( NMII ) , strain RSA493 clone 4 , was axenically grown in liquid ACCM-2 or ACCM-agarose at 37°C in 5% CO2 and 2 . 5% O2 as previously described [24] , [71] . When appropriate , kanamycin and chloramphenicol were added to ACCM-2 at 300 µg/ml and 3 µg/ml respectively . HeLa 229 cells ( CCL-2; ATCC , Manassas , VA ) and J774 . 1 cells were maintained in Dulbecco's Modified Eagle's Media ( DMEM ) supplemented with 10% heat inactivated fetal bovine serum ( FBS ) at 37°C in 5% CO2 . pKM225 was introduced into stationary phase C . burnetii NMII via electroporation at 18 kV , 500 Ω and 25 µF as previously described [28] , [71] . Following electroporation , the bacteria were recovered in 20 ml of ACCM-2 for 24 h before being plated on ACCM-agarose plates containing chloramphenicol . After 6 days incubation , single colonies were isolated and resuspended in 1 ml aliquots of ACCM-2 with chloramphenicol in 24 well plates . Following a further 6 days , each 1 ml C . burnetii culture was passaged 1∶1000 to provide bacteria for the vacuole formation assay and determination of the transposon insertion site . The remaining culture was pelleted via centrifugation and resuspended in 100 µl of DMEM containing 10% FBS and 10% Dimethyl sulfoxide for storage in 96 well plates . The genomic location of the transposon insert sites was determined for transposon mutants with distinct intracellular phenotypes and a wider random selection of recovered mutants . Nested primers within the transposon , facing the transposon-genome junction site , were used to amplify the insertion site either from C . burnetii cell lysate or purified genomic DNA . The first round of amplification used primer 1: GGGGGAAACGCCTGGTATC and a pool of random oligonucleotides with a common arm . The product of this PCR was used as a template for the second round of amplification with primer 2: GTCGGGTTTCGCCACCTC and primer ARB2: GGCCAGGCCTGCAGATGATG . The second PCR product was sequenced using primer 3: TCGATTTTTGTGATGCTCGTC . Sequencing results were analyzed using 4Peaks and BLAST programs . 1 . 5×104 HeLa 229 cells were added into 96 well tissue culture plates . The next day monolayers were infected with stationary phase C . burnetii transposon mutants at a multiplicity of infection ( MOI ) of approximately 500 in DMEM with 5% FBS . The infection was allowed to progress for approximately 96 h , with the media changed 24 h after infection . After 96 h , cells were fixed with 4% paraformaldehyde and then blocked and permeablized in blocking buffer ( PBS containing 2% BSA and 0 . 05% saponin ) . The cells were stained with anti-LAMP1 monoclonal H4A3 ( Developmental Studies Hybridoma Bank ) and rabbit anti-C . burnetii polyclonal antibody in blocking buffer at 1∶1000 and 1∶10000 respectively . Secondary antibodies , Alexa Fluor 488 and 594 ( Invitrogen ) were used at 1∶3000 also in blocking buffer . During final PBS washes bacterial and host DNA was stained with Hoechst 33342 ( Invitrogen ) . Stained infections were visually inspected for formation of large CCVs and those transposon mutants the exhibited abnormal CCV formation were investigated further . For additional immunofluorescence microscopy HeLa cells were added in 24-well dishes containing 10 mm glass coverslips . At the indicated times post-infection the cells were fixed and stained as above before being mounted on slides using ProLong Gold ( Invitrogen ) . For cathepsin D staining of infected cells , anti-cathepsin D ( Novus Biologicals ) was used at 1∶50 following fixation and permeablized with ice-cold MeOH and blocking in PBS with 2% BSA . For endogenous LC3 staining HeLa cells were infected for five days before fixing in cold methanol on ice for five minutes . Coverslips were blocked in 2%BSA , and stained with mouse anti-LC3 ( NanoTools clone 2G6 ) and rabbit anti-Coxiella antibodies at 1∶200 and 1∶10 , 000 , respectively , in blocking solution . Cells were washed three times in PBS and stained with anti-mouse 488 and anti-rabbit 546 at 1∶2000 . Coverslips were washed three times with PBS and mounted on slides using ProLong Gold Antifade reagent ( Invitrogen ) . Images for endogenous LC3 were acquired using an LSM510 confocal microscope equipped with a 100×/1 . 4 numerical aperture objective lens . Images were analyzed in Image J and Photoshop . For DQ Green BSA experiments , J774A . 1 cells were infected with C . burnetii NMII , cig2::Tn , or cig2::Tn pFLAG-Cig2 in 35 mm glass bottom dishes . Cells were incubated in medium containing 50 µg/ml DQ Green BSA at 36 h post-infection and allowed to incubate for a further 16 h . Cells were washed three times with PBS and placed in fresh 5% FBS/DMEM ( no phenol red ) as described previously [5] . Live images were acquired after one additional hour of incubation with the fresh media . Digital images were acquired with a Nikon Eclipse TE2000-S inverted fluorescence microscope using a 60×/1 . 4 or 100×/1 . 4 numerical aperture objective lens and a Photometrics CoolSNAP EZ camera controlled by SlideBook v . 5 . 5 imaging software . The day before infection , HeLa 229 were plated at a density of 5×104 into 24 well plates with or without 10 mm glass coverslips . Axenically grown stationary phase C . burnetii strains were quantified by qPCR using dotA specific primers [10] and diluted in DMEM with 5% FBS to an MOI of 50 . Following a 4 h infection period , cells were washed once with PBS and incubated with fresh DMEM with 5% FBS . This point was considered Day 0 and a sample was collected to provide the inoculum amount of C . burnetii . Infection lysate was collected at the time of infection , 24 ( Day 1 ) , 72 ( Day 3 ) , 120 ( Day 5 ) and 168 ( Day 7 ) h after this initial time point . Genomic DNA was extracted from these samples using the Illustra Bacteria GenomicPrep Mini Prep Kit ( GE Healthcare , Piscataway , NJ ) and was used to quantify genomic equivalents by dotA specific qPCR . In addition , replicate wells were fixed with 4% paraformaldehyde at Day 3 and Day 5 for subsequent immunofluorescent staining with anti-LAMP1 and anti-C . burnetii as described above . Translocation assays were performed as described previously [28] , [39] . Genes of interest were cloned into the SalI site of pJB-CAT-BlaM and these constructs were introduced into C . burnetii NMII via electroporation . BlaM fusion protein expression of isolated clones was confirmed by western blot with anti-BlaM ( 1∶1000 ) , ( QED Bioscience Inc , San Diego , CA ) . 2×104 HeLa cells were plated in black clear bottom 96 well trays and , the following day , were infected with stationary phase C . burnetii NMII strains at an MOI of 100 . The infection was allowed to proceed for 24 h before cells were loaded with the fluorescent substrate CCF4/AM according to the instructions for the LiveBLAzer-FRET B/G Loading Kit ( Invitrogen , Carlsbad , CA ) . Fluorescence , with an excitation of 415 nm and emission at 460 nm and 535 nm , was quantified using a Tecan M1000 plate reader . The ratio of signal at 460 nm to 535 nm ( blue:green ) was calculated relative to uninfected cells . In addition , cells were visualized by fluorescence microscopy using an inverted Nikon Eclipse TE-2000 S microscope and a 20× objective . In 24-well plates , HeLa229 cells were reverse-transfected with small-interfering RNA ( siRNA ) SMARTpools specific for human ATG5 ( NM_004849 ) , ATG12 ( NM_004707 ) , STX17 ( NM_017919 ) , or STX18 ( NM_016930 ) using Dharmafect-1 ( Thermo Scientific ) at a final concentration of 50 nM total siRNA in DMEM with 5% FBS . Transfected cells were incubated overnight , washed , and the adherent cells were subjected to a second round of siRNA transfection at the same concentration . After a two-day incubation , the cells were infected with C . burnetii NMII at a MOI of 50 . At one day post-infection , the cells were lifted and replated at a lower cell density into a 24-well plate containing 12-mm-diameter glass coverslips and incubated for an additional two days . Cells were processed for immunofluorescence as described above . Primers were designed to amplify ravZ or ravZC258A from plasmids pGFP-RavZ or pGFP-RavZC258A [60] by PCR and to contain extended overhangs specific for sequence- and ligation-independent cloning ( SLIC ) into the pJB-CAT-3×FLAG destination vector [72] . The following primers show the destination vector sequence underlined and the ravZ-specific sequences italicized: RavZ forward , 5′-ATATCGATTACAAGGATGACGATGACAAGGTCGACATGAAAGGCAAGTTAACAGG-3′ and RavZ reverse , 5′-GGGCGGGGCGTAAAAGCTTGCATGCCTCAGTCGACCTATTTTACCTTAATGCCACC-3′ . The resulting vectors pJB-CAT-3×FLAG-RavZ and pJB-CAT-3×FLAG-RavZC258A encode RavZ proteins that have three tandem copies of the FLAG epitope tag ( 3×FL ) fused to the amino terminus of the protein . Plasmid DNA ( pJB-CAT-3×FLAG-RavZ or pJB-CAT-3×FLAG-RavZ C258A ) was electroporated into C . burnetii NM II , and chloramphenicol-resistant C . burnetii were clonally isolated as described previously [28] . Immunoblots of C . burnetii lysates using anti-Flag M2 antibody ( Sigma ) confirmed that 3×FL-RavZ protein was expressed . C . burnetii transformed with pJB-CAT-3×FLAG , pJB-CAT-3×FLAG-RavZ or pJB-CAT-3×FLAG-RavZ C258A were used to infect HeLa cells at a MOI of 50 . After 10 h incubation , cells were washed and incubated for three days before either fixation with 4% PFA for immunofluorescence , or lysis for immunoblot analysis . Uninfected or three-day post-infection C . burnetii-infected HeLa cells were lysed as described previously for Figure 5 [60] . Lysates were centrifuged and the supernatant separated by SDS-PAGE for immunoblot analysis using an anti-LC3 antibody ( Novus ) at 1∶3000 and an anti-actin antibody ( Sigma ) at 1∶5000 . For Figure 6 , uninfected or five-day post infection C . burnetii infected HeLa cells were maintained in 6-well dishes before harvesting with a cell scraper and lysing in buffer containing 2% Triton-X as described in Tanida et al . , 2008 [73] . Cells were either left untreated prior to lysis , or were incubated in media containing 200 nM rapamycin and 100 ng/ml bafilomycin A1 2 h prior to lysis . | Coxiella burnetii is the causative agent of the human disease Q fever . This bacterium uses the Dot/Icm type IV secretion system to deliver effectors into the cytosol of host cells . The Dot/Icm system is required for intracellular replication of C . burnetii . To determine the contribution of individual proteins to the establishment of a vacuole that supports C . burnetii replication , we conducted a visual screen on a library of C . burnetii transposon insertion mutants and identified genes required for distinct stages of intracellular replication . This approach was validated through the identification of intracellular replication mutants that included insertions in most of the dot and icm genes , and through the identification of individual effector proteins delivered into host cell by the Dot/Icm system that participate in creating a vacuole that supports intracellular replication of C . burnetii . Complementation studies showed convincingly that the effector Cig57 was critical for intracellular replication . The effector protein Cig2 was found to play a unique role in promoting homotypic fusion of C . burnetii vacuoles . Disrupting host autophagy phenocopied the defect displayed by the cig2 mutant . Thus , our visual screen has successfully identified effectors required for intracellular replication of C . burnetii and indicates that Dot/Icm-dependent subversion of host autophagy promotes homotypic fusion of CCVs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"bacteriology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"host-pathogen",
"interactions",
"medical",
"microbiology",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"microbiology",
"pathogenesis",
"bacterial",
"pathogens"
] | 2014 | A Screen of Coxiella burnetii Mutants Reveals Important Roles for Dot/Icm Effectors and Host Autophagy in Vacuole Biogenesis |
Membrane proteins are critical functional molecules in the human body , constituting more than 30% of open reading frames in the human genome . Unfortunately , a myriad of difficulties in overexpression and reconstitution into membrane mimetics severely limit our ability to determine their structures . Computational tools are therefore instrumental to membrane protein structure prediction , consequently increasing our understanding of membrane protein function and their role in disease . Here , we describe a general framework facilitating membrane protein modeling and design that combines the scientific principles for membrane protein modeling with the flexible software architecture of Rosetta3 . This new framework , called RosettaMP , provides a general membrane representation that interfaces with scoring , conformational sampling , and mutation routines that can be easily combined to create new protocols . To demonstrate the capabilities of this implementation , we developed four proof-of-concept applications for ( 1 ) prediction of free energy changes upon mutation; ( 2 ) high-resolution structural refinement; ( 3 ) protein-protein docking; and ( 4 ) assembly of symmetric protein complexes , all in the membrane environment . Preliminary data show that these algorithms can produce meaningful scores and structures . The data also suggest needed improvements to both sampling routines and score functions . Importantly , the applications collectively demonstrate the potential of combining the flexible nature of RosettaMP with the power of Rosetta algorithms to facilitate membrane protein modeling and design .
Membrane proteins are critical participants in a wide variety of biological processes including cell adhesion , signaling , transport , and enzymatic activity [1] . They comprise more than 30% of open reading frames [2] and are targeted by over half of currently available pharmaceutical drugs [3 , 4] . Despite their importance , our knowledge of membrane protein structure and function remains severely limited , as shown by a constant 1–2% representation of structures in the Protein Data Bank [5] over the past decade [6] . The paucity of experimentally determined structures can be attributed to wide-ranging challenges in overexpression , reconstitution into membrane mimetics , and ultimately structure determination by various methods [7] . Due to these experimental challenges , computational approaches assume a pivotal role in advancing our understanding of membrane protein structure and function . Compared to modeling soluble proteins , membrane protein modeling has the advantage of constraining the conformational search space into the two dimensions of the membrane bilayer , which imposes structural constraints onto the protein . Whereas soluble proteins exhibit enormous structural diversity , the structural motifs in the membrane environment are either α-helical bundles or β-barrels . Since these folds are formed by secondary structure elements adopting preferred orientations in the ordered environment of the lipid bilayer , the use of adapted sampling techniques could substantially increase conformational sampling efficiency . A higher sampling efficiency is required because membrane proteins are typically much larger in size , offsetting the reduction in conformational search space . Additionally , computational methods for membrane protein modeling require reliable free energy calculations or score functions to distinguish native-like from non-native conformations . Therefore , an accurate representation of the heterogeneous environment of the lipid bilayer is needed . The membrane bilayer can be represented implicitly by using a layered , continuum solvation model , which is computationally inexpensive but unable to describe membrane fluctuations or specific membrane protein—lipid interactions . An additional challenge for the score function is that the precise location of the lipid bilayer surrounding the protein in experimental structures is unknown because the membrane mimetic evades experimental observation . The aforementioned challenges and the lack of experimental structures have delayed the development and therefore availability of high-quality computational methods for membrane protein modeling , compared to available methods for soluble protein modeling . Whereas soluble protein modeling increasingly focuses on high-resolution structural features as in docking , design and ligand docking applications , methods for membrane protein modeling still mainly focus on obtaining models for unknown protein structures . Four main techniques for computational modeling of membrane proteins are available: ( 1 ) Since template structures for homology modeling are unavailable for many membrane proteins of interest , ab initio modeling is an important technique ( e . g . using BCL∷MPFold [8–15] ) . Ab initio structure prediction is one of the most difficult of the modeling tasks , yet it also has the largest benefits because of its ability to predict novel folds . Additionally , in contrast to homology modeling where the final model can contain artifacts from the template , models from ab initio structure prediction are not biased by previously determined protein structures . ( 2 ) For low ( ~25% ) to very low ( ~5% ) sequence similarities to a known structure , fold recognition techniques generate a low-resolution protein model; the accuracy of these models rarely achieves better than 3–4 Å RMSD . ( 3 ) Homology modeling can be used to model the three-dimensional structure of a query protein if the sequence similarity between the query sequence and the sequence of a template structure is greater than ~30% . The recent increase in determined membrane protein structures ( and therefore template availability ) has elevated the quality and number of built homology models . Recently , GPCR homology models with an RMSD as low as 2 . 9 Å from the target structure were created from starting templates with a sequence identity as low as 15% [16] . ( 4 ) If the structure of the membrane protein is known , molecular dynamics ( MD ) simulations can follow time trajectories of proteins and lipids in full-atom representation with physics-based energy functions to investigate high-resolution phenomena such as ion channel gating or transport across the membrane [17–19] . With the recent increase in available membrane protein structures , high-resolution modeling methods including protein design have started to emerge [20–25] . Two notable achievements include a helix—helix interface design [21] and a design of a four-helix bundle that selectively transports metal ions across the membrane [20] . A limitation of many membrane protein modeling tools is high specialization to accomplish a single task; thus these methods are not easily combined with other modeling tools . The membrane protein community would benefit from an integrated tool that is able to carry out a variety of complex modeling tasks such as loop modeling , predicting the effects of mutations , design , docking , symmetric complex assembly , and ligand docking , in addition to ab initio structure prediction , homology modeling , and high-resolution refinement . Additionally , integrated methods would enable testing of a score function in multiple contexts to more rapidly converge on a universal score function in the bilayer environment . The Rosetta software suite offers an integrated toolset for biomolecular modeling , docking , and design , including a broadly tested and refined score function for soluble biomolecules . Moreover , Rosetta has two pioneering membrane protein modeling applications , RosettaMembrane ab initio [26] and relax [23] . The RosettaMembrane ab initio protocol [23 , 26] was one of the first methods for ab initio structure prediction of membrane proteins . It combines Rosetta’s ab initio structure prediction protocol for soluble proteins [27] with a low-resolution score function derived from a database of structures of membrane proteins [26 , 28 , 29] . This method was later updated to include a high-resolution refinement stage [23 , 30] that uses an all-atom score function based on the Lazaridis implicit Gaussian-exclusion solvation model for atoms in the membrane [31] . Recently , RosettaMembrane was also used to model transmembrane helical proteins from distant homologues [16] . Since the creation of the RosettaMembrane ab initio protocol in 2006 , Rosetta has been reorganized into a set of object-oriented libraries ( “Rosetta3” ) while RosettaMembrane remained in its original implementation . Rosetta3 is now a cohesive , flexible software suite that includes separate objects for conformation and scoring , interaction graphs , score functions organized by multi-body dependencies , kinematics managed through a fold tree , maps to identify flexible portions of the molecule ( s ) , job distribution , and scripting interfaces [32] . Continuous improvements and additions to Rosetta’s extensive library of complex tools motivate its use for membrane protein modeling . However , the scientific concepts of the original RosettaMembrane require integration and generalization to be compatible with the object-oriented architecture of Rosetta3 . Here , we present a new framework , called RosettaMP , integrated in the Rosetta software suite , which enables the development of novel protocols for membrane protein modeling and design . We describe the new , central building blocks to represent the membrane bilayer , and to sample and score both conformations and sequences . We used RosettaMP to create four proof-of-concept applications: ( 1 ) prediction of free energy changes upon mutation , ( 2 ) high-resolution structural refinement , ( 3 ) protein—protein docking , and ( 4 ) assembly of symmetric complexes , all in the membrane bilayer . The protocols can be accessed via command line , PyRosetta [33] , and RosettaScripts [34] , with various levels of customizability for both developers and users . Using a set of test cases , we are able to obtain information on the applicability of the existing score function in these wider contexts . Collectively , the applications demonstrate how RosettaMP and existing Rosetta protocols can be combined to quickly create powerful new methods to answer a broad range of scientific questions . Because of its ability to interoperate with existing Rosetta code and reusable representation of the lipid bilayer , RosettaMP substantially lowers the barrier in complexity for the development of new protocols to model and design membrane proteins , opening the door to many new , critically needed methods .
Here , we describe the use of a software suite for biomolecular modeling , docking , and design to enable rapid development of new applications targeted at membrane proteins , a class for which structure determination efforts are notoriously difficult . We used the scientific concepts in the RosettaMembrane structure prediction and refinement applications and created a modular framework within Rosetta3’s object-oriented architecture [32] . The four proof-of-concept applications demonstrate flexibility , generality , and simplicity of RosettaMP . The new framework enables combination of the membrane environment with a variety of Rosetta features: with the fold tree [36] , jumps in membrane proteins can be used to model multiple protein chains , flexible loops , and ligands; and with symmetry [64] , symmetric protein structure prediction , refinement and design will be feasible . RosettaMP will serve as the starting point for future protocol development , and each new application can be extensively tested and benchmarked . Our preliminary results show that RosettaMP has the potential to answer long-standing questions involving membrane proteins and lays the groundwork for the challenges that still remain . RosettaMP complements many existing tools for membrane protein modeling . MPrelax can be used to refine proteins inserted into the membrane using tools such as iMembrane [67] . MPrelax can also be used in combination with a contact prediction method to predict structures with low sequence similarities to their template ( similar to I-TASSER [68 , 69] ) . MPddG can directly be used for alanine scanning and extended for membrane protein- and interface-design . Homology models from MEDELLER [70] and Rosetta [16] can be used as input to MPdock and MPsymdock for modeling of large membrane protein complexes . In principle , the MPsymdock protocol can also be used to distinguish biological interfaces from non-native crystal contacts ( similar to COMP [71] or PISA for soluble proteins [72] ) . The variety of these potential applications shows that RosettaMP forms an important basis for new protocol development . The key components of any structure prediction or design algorithm are sampling and scoring . Conformational sampling routines are improved via RosettaMP through the connection of the membrane bilayer to the modeled biomolecule . This representation allows flexibility in choosing which object should be fixed vs . movable ( protein or membrane ) by representing the membrane as a ‘residue’ and using a jump in the fold tree . For example , a fixed bilayer enables sampling of membrane-embedded docking conformations in the MPdock and MPsymdock protocols , whereas a movable membrane decreases the computational cost of the MPrelax protocol . The latter allows optimizing the membrane position and orientation using minimization algorithms , resulting in lower scores for three of four cases . Moreover , the flexible linkage now permits constraining spans , chains , or proteins to the membrane in various depths and orientations , features that could not be modeled previously . The framework also simplifies implementation of enhanced sampling protocols through specialized movers that favor meaningful protein conformations in the membrane , for instance for ab initio prediction of α-helical and β-barrel membrane proteins using particular fragment types or favoring appropriate orientations and pairings of the secondary structure elements . For scoring , the new framework allows us to test Rosetta’s membrane score functions in new contexts . The four applications collectively demonstrate that the existing low- and high-resolution membrane protein score functions are generally effective , yet require further optimization . The preliminary MPddG application is able to identify favorable vs . disruptive mutations in the two tested cases and even produces a reasonable correlation for predicted vs . experimental ΔΔG values for an intra-membrane residue in OmpLA . Refinement , which is minimally modified from its original , tested implementation , captures the same minima . The naïve score functions for docking and symmetric docking exhibit minima for native-like interfaces in about half of the asymmetric and symmetric cases . These results are encouraging , especially since we have not made any changes to the RosettaMembrane score functions originally developed for folding and refinement . For future work , several improvements to the score function seem possible . Since the number of determined membrane protein structures has increased substantially in recent years , the low-resolution , knowledge-based score functions can be updated to reduce statistical errors . Further , existing score functions were solely derived from α-helical membrane proteins , and data from β-barrels could be used to create a distinct score function that could be tested with large-scale folding and refinement of these proteins . For instance , a recently derived hydrophobic potential for outer membrane β-barrels has been found to be condensed compared to that for α-helical membrane proteins , since bacterial outer membranes have a smaller membrane thickness [55 , 57] . An updated formulation of Rosetta’s distance-dependent dielectric electrostatic score [73] is needed to accommodate the low dielectric constant in the membrane . It is also now feasible in principle to sample protonated and deprotonated forms of ionizable residues with Rosetta pH [74 , 75]; however parameterization is needed to account for the insertion of charged species in the membrane . An advantage of RosettaMP is that new score functions can now be more easily derived than previously by using the score function machinery in Rosetta3 . The current membrane model is a flat bilayer model of fixed thickness , i . e . a slab model . RosettaMP could be a stepping-stone for tackling complex biological questions with more sophisticated membrane models . Effects needed to create the next generation of this model include intrinsic curvature , charge asymmetry [76–78] , and variable thickness [79] , attributed to the diverse repertoire of lipids that constitute the membrane environment [80] . Membrane thickness might play a role in the accurate estimation of ΔΔG values for interfacial aromatic residues [51] and in reproducing snorkeling of arginine and lysine to the membrane interface [54 , 81] . More sophisticated membrane models will be required for proteins that form pores or toroidal pores [82] . Another challenge is the modeling of membrane-anchored proteins or peptides [83] especially for small and/or unstructured peptides or half-helices inserting into the bilayer that are not identified as such by sequence-based prediction methods . RosettaMP will also enable the development of design protocols , an important yet challenging task with potential impact in synthetic biology and gene therapy . The latest membrane protein design efforts focused on helix-helix interfaces [20] , protein chimeras [84] , and used protein display [85] and combinatorial libraries [86] to identify promising designs . These efforts require manual processes and use full-atom energy functions derived mostly from MD force fields . With robust score functions and sampling routines , methods developed with RosettaMP will add to these emerging tools and complement MD simulation packages , enabling investigation of membrane protein structure , dynamics , and function from low- to high-resolution representations . In summary , we anticipate new progress by combining the power of existing Rosetta applications with RosettaMP . By making membrane protein modeling and design accessible to the broad scientific community , we hope to drive understanding of membrane protein structure , function and ultimately enable drug design for this essential class of proteins . | Over 30% of the human proteome consists of proteins embedded in biological membranes . These proteins are critical in many processes such as transport of materials in and out of the cell and transmitting signals to other cells in the body . They are implicated in a large number of diseases; in fact , they are targeted by over 50% of pharmaceutical drugs on the market . Since the membrane environment makes experimental structure determination extremely difficult , there is a need for alternative , computational approaches . Here , we describe a new framework , RosettaMP , for computational modeling and design of membrane protein structures , integrated in the Rosetta3 software suite . This framework includes a set of tools for representing the membrane bilayer , moving the protein , altering its sequence , and estimating free energies . We demonstrate tools to predict the effects of mutations , refine atomic details of protein structures , simulate protein binding , and assemble symmetric complexes , all in the membrane bilayer . Taken together , these applications demonstrate the potential of RosettaMP to facilitate membrane protein structure prediction and design , enabling us to understand the function of these proteins and their role in human disease . | [
"Abstract",
"Introduction",
"Discussion"
] | [] | 2015 | An Integrated Framework Advancing Membrane Protein Modeling and Design |
Colorectal cancer ( CRC ) is believed to arise from mutant stem cells in colonic crypts that undergo a well-characterized progression involving benign adenoma , the precursor to invasive carcinoma . Although a number of ( epi ) genetic events have been identified as drivers of this process , little is known about the dynamics involved in the stage-wise progression from the first appearance of an adenoma to its ultimate conversion to malignant cancer . By the time adenomas become endoscopically detectable ( i . e . , are in the range of 1–2 mm in diameter ) , adenomas are already comprised of hundreds of thousands of cells and may have been in existence for several years if not decades . Thus , a large fraction of adenomas may actually remain undetected during endoscopic screening and , at least in principle , could give rise to cancer before they are detected . It is therefore of importance to establish what fraction of adenomas is detectable , both as a function of when the colon is screened for neoplasia and as a function of the achievable detection limit . To this end , we have derived mathematical expressions for the detectable adenoma number and size distributions based on a recently developed stochastic model of CRC . Our results and illustrations using these expressions suggest ( 1 ) that screening efficacy is critically dependent on the detection threshold and implicit knowledge of the relevant stem cell fraction in adenomas , ( 2 ) that a large fraction of non-extinct adenomas remains likely undetected assuming plausible detection thresholds and cell division rates , and ( 3 ) , under a realistic description of adenoma initiation , growth and progression to CRC , the empirical prevalence of adenomas is likely inflated with lesions that are not on the pathway to cancer .
Adenomatous polyps ( or adenomas ) in the large intestine are considered benign precursors of colorectal cancer ( CRC ) and both clinical and molecular evidence suggest that they may sojourn for many years before turning into cancer [1] , [2] . For this reason , adenomas are considered a primary intervention target if detected and removed before they become malignant . However , questions remain regarding the significance of their histopathology , molecular signatures , as well as their number and sizes in average risk individuals . Since endoscopic screening for neoplastic lesions is generally limited by macroscopic detection thresholds ( of the order of a few mm in caliper size ) , a large fraction of adenomas may actually be missed , especially if the bulk of adenomas is too small for detection . Potentially , such “occult” adenomas could give rise to cancer before they are detected by endoscopy . Here we use a biologically-based model of colorectal carcinogenesis , which has previously been fitted to the age-specific incidence of CRC , to compute the number and size distributions of adenomas . Of particular interest is the fraction of detectable adenomas , as functions of age , detection threshold and the underlying cell kinetics in the adenomas . The underlying multistage clonal expansion ( MSCE ) model for CRC upon which our results are based explicitly considers the initiation , promotion and malignant conversion of adenomas [3]–[8] . According to this model , adenomas arise from normal colonic stem cells that suffer at least two rare rate-limiting events . We interpret these events as the biallelic inactivation of a tumor suppressor gene , in particular the APC tumor suppressor gene , which is the gene responsible for familial adenomatous polyposis ( FAP ) , and which is frequently mutated in colorectal neoplasia [9] . The inactivation of APC is understood to occur in colonic crypts ( the fundamental proliferative unit in the colon ) whose stem cells have previously acquired a mutation at one of the two APC alleles . Because the process of adenoma formation may involve additional genes ( such as KRAS ) , we extend the model framework to accommodate additional rate-limiting mutations for the initiation of an adenoma and generalize the mathematical derivation of their number and size distribution accordingly . However , there is both clinical and experimental evidence that the number of requisite rate-limiting events or mutations for adenoma initiation is small . Once a stem cell is initiated in this model , it is free to proliferate . The basic version of our CRC model assumes that adenoma initiation occurs when the remaining wild-type copy of the APC tumor suppressor gene is deleted or mutated in a stem cell of a ( pre-initiated ) APC+/− colonic crypt . In a more realistic model , which is supported by recent experimental findings in the murine system [10] , we also model the transient amplification of APC−/− stem cells prior to their clonal expansion , effectively adding a stage to the initiation process [10]–[12] . The theoretical results derived here are complemented by model predictions for the adenoma size distributions and their ( age-specific ) prevalence based on parameter estimates obtained previously from fitting cancer incidence data . Since not all biological model parameters can be directly estimated from incidence data alone ( non-identifiability issue ) , we explore the sensitivity of our findings by varying unknown parameters , such as the cell division rate of initiated stem cells , within their plausible ranges . In spite of the model uncertainties and the lack of precise clinical data on adenoma number and sizes , a biologically based approach that is broadly consistent with the pathogenesis of CRC makes it possible to explore more rationally the impact of risk factors and interventions on adenoma development and cancer progression .
First we briefly review the MSCE model for CRC and then introduce the notation for the relevant stochastic processes involved in the formation of adenomas . We have previously derived expressions for the number and size distribution of non-extinct pre-malignant clones in the context of the two-stage clonal expansion ( TSCE ) model [13] . This model assumes that the clones develop from a ( deterministic ) source of progenitor cells via a non-homogeneous Poisson process . An important extension of this work was put forward by Dewanji et al . [14] for the size distribution of a random sum of Poisson-generated ( pre-malignant ) clones , which corresponds to a generalized Luria-Delbrück ( GLD ) distribution for mutant colonies . A hallmark of this distribution is a long tail reflecting large fluctuations of the total ( mutant ) population size . A further extension derived expressions for the number and size distributions of pre-malignant clones conditioning on observations from individuals who have not previously been diagnosed with CRC [6] . For we have only one pre-initiation event , defined by a -mutation , and -cells are the initiated cells . As mentioned in the previous section , the -mutation follows a PP formulation . Let denote the size of the adenoma at time with the first pre-initiation ( -mutation ) time . The distribution of is given by the GLD distribution previously derived by Dewanji et al . [14] , for the process originating at time and involving the initiation rate and the birth and death rates of the initiated cells given by and , respectively . As mentioned before , we assume that the generation of -cells follows a non-homogeneous PP with rate , where gives the deterministic growth curve for the normal stem cells in the tissue . Let be the number of first pre-initiations ( -mutations ) by time and let be the occurrence times of these -mutations . Also , write , where is the indicator function . That is , , if the corresponding adenoma is detectable at time , and 0 otherwise . Then , the number of detectable adenoma can be written as a filtered Poisson processThe probability generating function ( PGF ) of can be written as ( 6 ) where is the PGF of the binary variate with success probabilityNote that is the probability that a -mutation taking place at time results in a detectable adenoma at time . This probability can be obtained from the distribution of . For constant parameters , this reduces to , using ( 1 ) with , . Here the two pre-initiation events ( - and -mutations ) precede initiation and growth of initiated -cells into sub-clones . Let denote the size of the adenoma at time with the corresponding - and -mutations taking place at times and , respectively . Note that the distribution of is given by the GLD distribution originating at time and involving the initiation rate and the birth and death rates of the initiated or -cells , given by and , respectively . We derive explicit expressions for the number and size distributions for the case when both - and -mutations are of PP-type . The case when the -mutations are of AD-type is described in the online supplement ( Text S1 ) . As before , the number of detectable adenomas at time can be written as ( 15 ) where is the number of detectable adenomas that emerged from a -cell born at time . Then , as in ( 6 ) , the PGF of can be expressed by ( 16 ) where is the PGF of . Using the Lemma and eq . ( 10 ) of Dewanji et al . [14] , we have further ( 17 ) and for , ( 18 ) where , for . Again , ( 19 ) where this sum is over all the -mutations by time that occurred at times and which emerged from a -cell that was born at time . Here , if the adenoma originated from the -cell born at time and the -cell born at time is detectable at time and 0 otherwise; that is , . Note that , since does not depend on . Therefore , as in the previous section , is a filtered PP with the PGF , similar to that in ( 6 ) . Also , is a non-homogeneous PP with rate , for , and for fixed , is a Poisson variate with mean , where ( 20 ) This probability can be obtained from the distribution of given in ( 1 ) . The notation introduced in the previous section is easily generalized to . The random variable denotes the size of the adenoma at time with the corresponding - , , -mutation times , respectively . The distribution of is , as before , given by the GLD distribution with time origin at and with initiation rate . Initiated -cells divide or die with rates and , respectively . Importantly , depends on alone ( i . e . , ) , and the distribution is given by ( 1 ) for constant parameters . Various combinations of AD-type and PP-type generations for the pre-initiations are possible , but the formulation of the likelihood becomes more complicated . The special cases when all pre-initiations are of PP-type or AD-type are covered in the online supplement ( Text S1 ) .
The derived expressions allow us to readily predict both observable and unobservable numbers of pre-malignant tumors in a tissue . Such predictions are helpful in validating cancer models using intermediate endpoints on precursor lesions , in particular their number and sizes . Furthermore , being able to predict the unobserved portion of precursor lesions is of clinical relevance for early detection and cancer prevention . Here , we illustrate the utility of the derived expressions using the example of colonic adenomas . Specifically , we present the predicted size distribution of adenomas and the age-specific adenoma prevalence , i . e . , the probability of finding at least one observable adenoma in an individual as a function of age . Since population-level screening is typically performed on asymptomatic individuals , we also condition on individuals having not developed cancer in the tissue of interest at the time of observation . The predictions presented here are for as described above . The underlying CRC model for cancer incidence is the 4-stage model previously derived by Luebeck & Moolgavkar [12] and updated by Meza et al . [7] . The alternative model ( PP for - and AD for -mutation for in the online supplement ( Text S1 ) ) yields very similar results ( not shown ) . Importantly , not all biological parameters of the MSCE/CRC model are estimable from incidence data alone . For example , for the 4-stage model used here , only the parameters , the product ( slope parameter ) , and the ratio are identifiable . However , if the cell division rate of initiated cells , , is known , all parameters of the model can be determined ( assuming that the number of normal tissue stem cells , , is known and that ) . For our illustrations , we choose plausible values for the cell division rate , but keep the values of , and as estimated by Meza et al . [7] . This affords explicit computation of the adenoma number and size distribution without altering the fits of the model to the observed CRC incidence . Figure 2 ( left panel ) shows the predicted size distribution of non-extinct adenoma without an imposed detection threshold ( i . e . , ) for the model with . With constant parameters , both the unconditional and conditional ( on no prior CRC development ) size distributions of detectable adenoma are given by expressions ( 24 ) and ( 31 ) , respectively . For sizes sufficiently large , the unconditional adenoma size distribution is roughly log-log-linear , while the conditional size distribution shows departures from this behavior for sizes above cells , i . e . , when the risk of an adenoma-to-carcinoma transition increases more rapidly . This phenomenon is more noticeable when the cell division rate is lower . Figure 2 ( right panel ) shows the probability of detecting an adenoma at age 70 as a function of the detection threshold for both unconditional ( solid ) and conditional ( dashed ) adenoma size distributions . Higher cell division rates ( ) give rise to larger adenoma sizes and hence lead to higher detection probabilities even though the net cell proliferation rate ( ) is approximately the same . For constant parameters , the unconditional and conditional detection probabilities are given by ( 23 ) and ( 30 ) , respectively . This figure reveals that even for relatively small ( i . e . , sensitive ) thresholds of a few thousand cells , many adenomas may go undetected . However , the precise proportion of detectable adenomas depends on the cell division rate with higher values of making detection more likely . Figure 3 shows the predicted age-specific adenoma prevalence in asymptomatic individuals for both males and females and for the models with and , as described by Meza et al . [7] , including their dependence on the observation threshold . The prevalence is defined as the probability of at least one detectable adenoma at age , and is given by ( 11 ) for and ( 29 ) for . In comparison with careful observations from autopsy studies [16] , the model under-estimates these empirical data ( represented by filled circles ) unless one accepts a very small number of initiated stem cells to be observable . There are several explanations why the model-generated expected prevalence of adenoma might be lower than the clinical data would indicate ( see the end of Discussion ) .
We have previously derived number and size distributions of pre-malignant clones in the context of the two-stage clonal expansion ( TSCE ) model of carcinogenesis [13] , [17] and more recently established a formal connection of these results with fluctuation analyses based on the Luria-Delbrück distribution [14] . The mathematical tools derived in these papers were subsequently applied to the problem of screening for colorectal adenoma allowing for interventions resulting from their complete or incomplete removal [6] . These explorations , however , required time consuming computer simulations . In contrast , here we derive mathematical expressions that allow us to readily compute adenoma number and size distributions without simulation . These expressions can form the basis for computing the likelihood of adenoma data from screening studies involving sigmoidoscopies , colonoscopies and computed-tomographic colonographies , and thus are of significantly practical importance . Moreover the analytical form of the likelihood function allows for parameter estimation and likelihood-based hypothesis testing . Analyses of such data will be forthcoming and are the subject of a separate paper . Our previous analyses of CRC incidence data suggest that , the number of requisite pre-initiation mutations , is indeed small [7] , [12] . corresponds to a ‘two-hit’ model for initiation , in essence representing the biallelic inactivation of a tumor suppressor gene ( Knudson's recessive oncogenesis ) [7] . A model with may describe both the inactivation of a tumor suppressor gene as well as the activation of an oncogene [7] , [12] . Here , we also treat the case of general , which can be viewed as a model for clonal evolution due to the tree-structure where the nodes represent immortal mutant stem cells that will give rise to specific sub-clones which may or may not be identified as such . We distinguish between two types of rate-limiting events , one that generates ( potentially multiple ) mutations via asymmetric cell division while preserving the progenitor stem cell from which the mutations arose ( PP-type ) , and one that leads to a transition of a progenitor cell into one cell that acquires a new mutation ( AD-type ) . The MSCE model used here assumes that all events that lead to initiated cells are PP-type . This is only a mild restriction since for rare events the PP-type emission is equivalent to a AD-type transition ( see Figure 1 ) . For frequent ( high rate ) events , the AD-type transition looses its rate-limiting nature and can be ignored , while the high rate ( PP-type ) process leads to the accumulation of multiple clones and thus has the potential to capture non-mutational events , such as the transient amplification of proliferative cells from resident stem cells in the colonic crypts . Once a stem cell is considered initiated , i . e . , is of type , we assume that it undergoes a stochastic birth-death process . This leads to the GLD distribution introduced in [14] for the adenoma size , which reduces to a Negative Binomial distribution for constant parameters . Note , however , our formalism is more general and can accommodate other growth models that do not result in a GLD size distribution for the initiated cell population emerging from a progenitor cell [17] , [18] . Finally , we wish to comment on the predictions of the model for the age-specific adenoma prevalence ( Figure 3 ) . In comparison with the empirical data , our predictions appear too low . However , the discrepancy depends on what is assumed for the initiated stem cell division rate and the detection threshold . While the range of plausible values for is limited by how fast initiated stem cell can cycle in the adenoma ( unlikely more than 2–3 times a week ) , it is not clear what fraction of cells in an adenoma is truly at risk for malignant transformation [10] . Assuming that a 1 mm adenoma , the caliper size detection limit cited by Clark et al . [16] , contains about 500 , 000 cells [19] and that only 1–10% of cells in an adenoma are tumor stem cells [10] , [20] , may be as low as 5000 cells and therefore the discrepancy may be less dramatic . Alternatively , one might include pre-initiated cells in the adenoma size count . However , our assumption is that pre-initiated cells do not expand clonally , although they may increase in number as a result of multiple births of the same type of mutation from a single stem cell over time ( via Poisson process emissions ) . Thus , since locus-specific mutations are rare ( of the order of to per year ) , the contribution of pre-initiated cells to the overall number of cells in an adenoma is likely very small . It is well-recognized that adenomas can be genetically diverse and differ widely in their neoplastic potential . Indeed , adenomas have been suggested to regress implying that there are adenomas that are not on the pathway to cancer [21] , although regression may simply reflect the stochastic nature of tumor growth . A more intriguing possibility of resolving the discrepancy is that adenomas go through a growth-bottleneck ( i . e . , stagnancy ) before they can become cancerous . In this scenario , adenomas might sojourn in a reservoir until an activating mutation or change in tumor microenvironment releases them from arrest [22] , [23] . Although incorporating this scenario into the MSCE model may be challenging , the framework presented here is independent of the particular dynamics of the initiated cells and the number of clonal expansions assumed . | The adenomatous polyp ( or adenoma ) is considered the common precursor lesion for colorectal cancer ( CRC ) . Although the natural history of adenomas is well-characterized in terms of their histopathology and ( epi ) genomic changes , little is known about their dynamics in the stage-wise progression from the first appearance of an adenoma to its conversion to malignant cancer . By the time adenomas become endoscopically detectable ( i . e . , are in the range of 1–2 mm in diameter ) , adenomas are already comprised of hundreds of thousands of cells . A large fraction of adenomas may therefore remain undetected during screening and , in spite of their small ( subthreshold ) size , could give rise to cancer prior to being detected . It is therefore of importance to establish what fraction of adenomas is detectable , both as a function of the age at screening for colorectal neoplasia and the size ( threshold ) above which adenomas can be detected reliably . Here we derive mathematical expressions for the distribution of adenoma number and sizes based on a recently developed stochastic model for CRC , which has previously been calibrated and validated against age-specific CRC incidence data . | [
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion"
] | [
"oncology",
"medicine",
"cancer",
"detection",
"and",
"diagnosis",
"stochastic",
"processes",
"mathematics",
"cancer",
"prevention",
"cancers",
"and",
"neoplasms",
"probability",
"theory"
] | 2011 | Number and Size Distribution of Colorectal Adenomas under the Multistage Clonal Expansion Model of Cancer |
Whooping cough caused by Bordetella pertussis is a re-emerging infectious disease despite the introduction of safer acellular pertussis vaccines ( Pa ) . One explanation for this is that Pa are less protective than the more reactogenic whole cell pertussis vaccines ( Pw ) that they replaced . Although Pa induce potent antibody responses , and protection has been found to be associated with high concentrations of circulating IgG against vaccine antigens , it has not been firmly established that host protection induced with this vaccine is mediated solely by humoral immunity . The aim of this study was to examine the relative contribution of Th1 and Th17 cells in host immunity to infection with B . pertussis and in immunity induced by immunization with Pw and Pa and to use this information to help rationally design a more effective Pa . Our findings demonstrate that Th1 and Th17 both function in protective immunity induced by infection with B . pertussis or immunization with Pw . In contrast , a current licensed Pa , administered with alum as the adjuvant , induced Th2 and Th17 cells , but weak Th1 responses . We found that IL-1 signalling played a central role in protective immunity induced with alum-adsorbed Pa and this was associated with the induction of Th17 cells . Pa generated strong antibody and Th2 responses , but was fully protective in IL-4-defective mice , suggesting that Th2 cells were dispensable . In contrast , Pa failed to confer protective immunity in IL-17A-defective mice . Bacterial clearance mediated by Pa-induced Th17 cells was associated with cell recruitment to the lungs after challenge . Finally , protective immunity induced by an experimental Pa could be enhanced by substituting alum with a TLR agonist that induces Th1 cells . Our findings demonstrate that alum promotes protective immunity through IL-1β-induced IL-17A production , but also reveal that optimum protection against B . pertussis requires induction of Th1 , but not Th2 cells .
Bordetella pertussis is a Gram-negative bacterium that causes whooping cough ( pertussis ) , a severe respiratory tract infection that kills almost 200 , 000 children annually worldwide . Whole cell vaccines ( Pw ) introduced in the 1950s significantly reduced the incidence of pertussis but were associated with side effects and were replaced by safer acellular pertussis vaccines ( Pa ) in most developed countries following successful clinical trials in the 1990s [1]–[3] . However the incidence of pertussis is increasing , especially in adolescents and adults [4] , [5] and this may be related to suboptimal or waning immunity induced by Pa [6] . Despite recent progress , the mechanism of protective immunity induced by pertussis vaccines remains unclear . Analysis of serological responses in immunized children revealed a correlation between antibody response to the B . pertussis antigens , pertactin , pertussis toxin ( PT ) or fimbrae and Pa-induced protection [7] . Analysis of T cell responses in children demonstrated that Pa promote Th2-type responses , whereas Pw preferentially induce Th1 cells [8] , [9] . Studies in mouse models have suggested that Th1 cells play a critical role in immunity induced by Pw or previous infection , whereas Th2 cells and antibody confer protection induced by Pa [10]–[13] . However it has also been reported that the superior long term protection induced by Pw in mice , when antibody responses had waned significantly , was associated with the induction of potent Th1 responses [14] . More recently it has been reported that Th17 cells also play a role in protection induced by natural infection or immunization with Pw [15]–[18] , but their role in Pa-induced immunity has not been examined . Like most other licensed infectious disease vaccines , Pa are delivered to children using alum as the adjuvant . Traditionally it had been accepted that alum enhances immune responses to the antigens in a vaccine by facilitating retention of the antigen at the site of injection , thus promoting antibody responses and antigen uptake by antigen presenting cells for priming of T cell responses in the draining lymph nodes [19] . It also emerged that alum preferentially promoted Th2 cells , which are considered to be important for protection against parasites and extracellular bacteria by providing help for antibody production . More recently , it was demonstrated that alum functions as an adjuvant in mice by activating the Nlrp3 inflammasome [20] , [21] , involved in processing of IL-1β . It has also been reported that activation of caspase-1 and Nlrp3 , although required for IL-1β production , were dispensable for alum-mediated Th2-associated antibody production [22] . However , the role of Th17 cells has not been addressed . We and others have shown that caspase-1-processed IL-1β plays a crucial role in the induction of Th17 cells that mediate autoimmunity [23]–[25] . Th17 cells are also required for protective immunity against infection , primarily fungi and extracellular bacteria , such as Klebsiella pneumonia , where IL-17 promotes recruitment of neutrophils [26] . The aim of this study was to examine the relative contribution of Th1 and Th17 cells in host immunity to B . pertussis , both in the clearance of a primary infection in naive mice and in response to vaccination and to use this information to help in the rational design of a more effective Pa . Our findings demonstrate both Th1 and Th17 cells contribute to clearance of a primary infection of mice with B . pertussis , and that IFN-γ has a critical role in adaptive immunity to B . pertussis induced by Pw . In contrast , an alum-adjuvanted Pa induced Th17 as well as Th2-type responses , but surprisingly we found that IL-17A played an essential role , while IL-4 was unnecessary for bacterial clearance . The induction of Th17 responses by Pa required activation of IL-1R-signalling in innate immune cells and protection was associated with cellular recruitment to the lungs after challenge with B . pertussis and activation of bacterial killing by neutrophils . Furthermore , the protective efficacy of experimental Pa could be enhanced to that of Pw by substituting alum with an adjuvant that induces Th1 cells .
Previous infection with B . pertussis is effective in inducing protective immunity against subsequent infection and this has been associated with the induction of Th1 cells [13] , [27] . Indeed , it has already been established that IFN-γ plays a critical role in clearance of a primary infection with B . pertussis [10] , [11] . However there is also evidence that Th17 cells may be involved [16] , [18] . Here we examined the relative role of T cell subtypes in host immunity to a primary infection with B . pertussis in naive mice and first concentrated on defining the role of IL-17 . We found that infection of mice with B . pertussis was associated with induction of B . pertussis-specific Th17 cells . Antigen-specific IL-17A ( Figure 1A ) and IL-17F ( Figure S1A ) production was detected in lungs as early as 7 days post challenge and reached a peak after 3–4 weeks . Interestingly , B . pertussis filamentous hemagglutinin ( FHA ) , which is considered to be the least important antigen in Pa from the perspective of antibody responses [7] , was a major target for Th17 cells from infected mice ( Figure S1B ) . In order to confirm these findings and to examine the cellular source of IL-17 , we performed intracellular cytokine staining ( ICS ) and flow cytometry analysis on lung mononuclear cells ex vivo , without re-stimulation . We found significant increase in the frequency ( Figure 1B , C ) and absolute numbers ( Figure S2 ) of IL-17A-producing CD4 T cells in the lung throughout the course of infection with B . pertussis . The earlier peak of IL-17A+CD4+ T cells ( day 14 ) compared with antigen-specific IL-17A detected by ELISA ( day 21 ) , probably reflect the difference in the assay system , with the latter involving a re-stimulation in vitro and therefore including memory cells , while the ICS was a more direct ex vivo measure of activated effector Th17 cells . Taken together these data show that B . pertussis infection of mice induces significant numbers of B . pertussis-specific Th17 cells in the lungs . In order to examine the role of IL-17A in bacterial clearance , we compared the course of infection in IL-17A-defective ( IL-17A−/− ) and WT mice . IL-17A−/− mice had 100–1000 fold more CFU in the lungs at the later stages of infection with bacteria still detectable in the lungs up to week 6 ( Figure 1D ) . The more severe infection in IL-17A−/− mice was associated with a significant reduction in CXCL1 ( KC ) production ( Figure S3 ) and impaired neutrophil recruitment ( Figure 1E ) to the lungs post challenge . We used an adoptive cell transfer approach to examine the relative role of Th1 and Th17 cells in protective immunity to B . pertussis . We generated polarized B . pertussis-specific Th1 or Th17 cells ( Figure S4 ) by culture of spleen cells from convalescent WT or IFN-γ−/− ( to overcome the problems of reversion of Th17 cells to Th1 ) with antigen and IL-12 or IL-1β and IL-23 respectively . Transfer of either Th1 or Th17 cells alone before B . pertussis challenge reduced the CFU counts by about 10 fold over the course of infection ( Figure 1F ) . Transfer of both populations together had a greater effect with CFU count significantly reduced by 50–100 fold compared to controls . In contrast transfer of naïve T cells from WT mice failed to confer protection to infected mice . These findings demonstrate that both Th1 and Th17 cells contribute to natural immunity induced by infection with B . pertussis in mice . Pw are more protective than Pa in mice [12] , [13] , [28] , [29] , which is even more pronounced when mice are challenged at an extended interval after immunization [14] and this has been attributed to the induction of Th1 cells by Pw [13] . Although a Connaught laboratories Pw only had an efficacy of 36 or 48% compared with 84 or 85% for 3 and 5-component Pa in the pertussis clinical trials carried out in Sweden and Italy in the 1990s [1] , [2] , most good Pw have efficacy of 93–96% in children [3] , [30] , [31] and a UK Pw was significantly more protective than the three-component Pa in a randomized controlled trial [32] . Here we examined the relative roles of IFN-γ and IL-17 in clearance of B . pertussis from the respiratory tract of mice immunized with Pw . We used a plain ( without alum ) Pw reference preparation . Although most recent Pw are absorbed to alum , plain Pw , such as the one manufactured by Wellcome laboratories , were routinely used until the 1980s in many European countries and had high efficacy against pertussis [30] , [33] . Furthermore , we have found that plain Pw induce similar immune responses and protection against infection as alum-absorbed Pw [12] [and Mills , unpublished] . Here we found that protective immunity induced by Pw was significantly compromised in IFN-γ−/− mice , with 100–1000 fold more bacteria in the lungs compared with Pw-immunized WT mice 3 , 7 and 10 days after aerosol challenge ( Figure 2A ) . The CFU counts were also significantly higher in Pw-immunized IL-17A−/− compared with WT mice 3 days post B . pertussis aerosol challenge , but IL-17A−/− mice , like WT mice had cleared the infection by day 7 . Mice immunized with Pw developed strong Th1 responses with high concentrations of IFN-γ produced by spleen cells from Pw immunized WT and IL-17A−/− mice , which was undetectable in IFN-γ−/− mice ( Figure 2B ) . B . pertussis-specific IL-17 was also induced by Pw and this was enhanced in IFN-γ−/− mice . B . pertussis-specific IL-13 was at background concentrations in spleen cells from Pw-immunized WT mice , but was induced at significant concentrations in IFN-γ−/− mice ( Figure 2B ) . These findings demonstrate that Pw induce Th1 and Th17 cells and confer protective immunity in mice via IFN-γ induction , but that IL-17A also contributes , though less significantly . Having shown that protection induced by Pw is mediated largely by Th1 cells , we examined the mechanism of host immunity induced by immunization with a licensed alum-absorbed Pa . Immunization with Pa by either i . p . or i . m . routes conferred protection against B . pertussis infection ( Figure S5A ) . We have previously reported that Pa selectively induced Th2-type responses whereas Pw promoted Th1 responses [12] , [13] . Here we found that Pa also induced B . pertussis-specific IL-17A from CD4+ T cells ( Figure S5B ) . We next examined the role of Th17 versus Th1 and Th2 cells in Pa-induced immunity . The bacterial clearance curves were almost identical for Pa-immunized WT and IL-4−/− or IFN-γ−/− mice ( Figure 3A ) . In contrast , the rate of bacterial clearance was dramatically slower in IL-17A−/− mice , with 100 fold more bacteria on day 3 and significant bacteria in the lungs on day 10 , when the WT mice had cleared the infection ( Figure 3A ) . Pa still induced Th2 responses in IL-17A−/− mice , with B . pertussis-specific IL-13 similar to that in WT mice ( Figure 3B ) . In contrast , B . pertussis-specific IL-13 production by spleen cells was close to background concentrations in IL-4−/− mice , whereas IL-17 was similar to that seen in Pa-immunized WT mice . FHA-specific IFN-γ was undetectable in Pa immunized mice and the low levels of IFN-γ detected in response to HKBp was not significantly different between WT , IL-17A−/− and IL-4−/− mice ( Figure 3B ) . Collectively these findings demonstrate an essential role for IL-17A , but not for IL-4 or IFN-γ , in protective immunity induced by Pa in mice . An examination of antibody responses revealed that total IgG and IgG1 were significantly reduced in both IL-4−/− and IL-17A−/− mice ( Figure 3C ) . IgG2a ( Figure 3C ) and IgG2c ( data not shown ) were significantly higher in IL-4−/− than WT mice , but similar in IL-17A−/− and WT mice . To examine the mechanism of immune protection mediated by IL-17A in the lungs , we investigated phagocytic cell influx upon B . pertussis challenge . There was a significant increase in the recruitment of both neutrophils and macrophages to the lungs after B . pertussis challenge in Pa-immunized mice compared to non-immunized mice , which peaked at day 7 post challenge ( Figure 4A ) . Cellular recruitment to the lungs was similar in Pa-immunized WT and IL-4−/− mice . In contrast , the influx of neutrophils and macrophages was significantly reduced in Pa-immunized IL-17A−/− mice . This was associated with dramatically lower CXCL1 and CCL3 ( MIP-1α ) concentrations in the lungs of Pa-immunized IL-17A−/− compared with WT or IL-4−/− mice post challenge ( Figure 4B ) . We have previously reported that IL-17 can promote macrophage killing of B . pertussis [15] . Here we demonstrate that neutrophils were also capable of killing B . pertussis following opsonisation with normal mouse serum , with killing detected after 1–3 hours , and this was significantly enhanced by IL-17A or IFN-γ but not IL-17F ( Figure 4C ) . Furthermore , killing was further , though not significantly , enhanced following addition of immune serum from Pa-immunized mice ( Figure 4C ) . These findings suggest that Pa-induced IL-17A enhances chemokine production , which recruits macrophages and neutrophils to the lungs soon after challenge with B . pertussis and these cells mediate killing of B . pertussis . Contrary to the perceived wisdom , our study suggests that Th2 cells are unnecessary , and Th17 cells play a critical role in protective immunity induced with Pa . We examined the mechanism of Th17 cell induction with Pa , in particular the role of Nlrp3 and IL-1 . It had previously been reported that alum functions as an adjuvant by activation of the Nlrp3 inflammasome [20] , [21] , although this has been questioned by others [22] . Furthermore , we had previously shown that IL-1β , induced via caspase-1 and Nlrp3 , plays a critical role in IL-17-mediated pathology in autoimmune disease [23] , [25] , [34] . Here we found that Pa or alum induced significant concentrations of IL-1β from LPS-primed DC and this was significantly reduced following addition of a caspase-1 inhibitor ( Figure S6A ) or using DC from Nlrp3−/− mice ( Figure S6B ) . Furthermore , significant concentrations of IL-1β were detected in draining lymph nodes 4 hours after injection of Pa ( Figure S6C ) . These finding demonstrate that alum-absorbed Pa promotes IL-1β production by DC in vitro via activation of caspase-1 and Nlrp3 . We also examined the role of Nlrp3 in the protective efficacy of Pa in vivo . Bacterial clearance was reduced though only significantly on day 5 post challenge in Pa-immunized Nlrp3−/− compared with WT mice ( Figure S7A ) . Furthermore , IL-17A production determined by ELISA on B . pertussis antigen-stimulated spleen cells ( Figure S7B ) or by intracellular cytokine staining on CD4+ T cells ( Figure S7C ) was similar in Pa-immunized Nlrp3−/− and WT mice . Finally IL-1β production in the lungs of B . pertussis infected mice was not significantly different between Nlrp3−/− and WT mice ( Figure S7D ) . In contrast to the rather limited attenuation of anti-B . pertussis immunity in Pa-immunized Nlrp3−/− mice , we found a dramatic reduction in the rate of bacterial clearance in Pa-immunized IL-1RI−/− mice , with 1000 fold more bacteria in the lungs when compared with Pa-immunized WT mice at 3 and 7 days post challenge ( Figure 5A ) . Furthermore , WT mice had completely cleared the bacteria by day 10 , where as there were significant numbers of bacteria in the lungs of IL-1RI−/− mice at this time point . These findings demonstrate that IL-1 is critical for protection , and its induction in vitro is dependent on caspase-1 and NLRP3 , but in vivo NLRP3 appears to be dispensable , suggesting that NLPR3-independent IL-1 pathways may be involved . Immunization of WT mice with Pa induced strong Th2-type responses and Th17 responses . However B . pertussis-specific IL-17A production was undetectable in spleen cells from IL-1RI−/− mice immunized with Pa ( Figure 5B ) . In contrast , IL-4 and IL-13 production was similar in Pa-immunized WT and IL-1RI−/− mice . In addition , B . pertussis-specific IL-17 was detectable in lungs 7 and 10 days after challenge of WT mice immunized with Pa , but was completely undetectable in IL-1RI−/− mice ( Figure 5C ) . In contrast , significant concentrations of IL-13 were detected in the lungs 7 and 10 days after B . pertussis challenge of WT and IL-1RI−/− mice immunized with Pa ( Figure 5C ) . Immunization of mice with Pa induced potent antibody responses , predominantly of the IgG1 subclass; this was significantly reduced in IL-1RI−/− mice ( Figure 5D ) . Pa generated weaker IgG2a and IgG2c antibody responses , which were also reduced in IL-1RI−/− mice . These findings demonstrate that IL-1 signalling plays an essential role in Pa-induced immunity and this involves induction of Th17 cells and antibody , but not Th2 cells . Since previous infection , and immunization with Pw induce potent Th1 cells and confer high levels of protection against B . pertussis [12]–[14] , [28] , [29] ( and present study ) , we examined the hypothesis that the efficacy of Pa could be enhanced by substituting alum with an adjuvant such as CpG , which promotes Th1 cells [35] . Since commercially available vaccines are already adsorbed to alum , here we used a laboratory prepared vaccine composed of the two B . pertussis antigens used in all licensed Pa , FHA and detoxified PT . Immunization of mice with the antigens ( Ag ) alone without adjuvant failed to confer immunity against B . pertussis infection ( Figure 6A ) . Consistent with our earlier studies , Ag formulated with alum conferred a good level of protection , however , this was significantly enhanced when the Ag were formulated with CpG . Bacteria were undetectable on days 10 and 14 post challenge in mice immunized with Ag and CpG , but were still detectable in mice immunized with Ag and alum ( Figure 6A ) . An examination of immune responses on the day of challenge revealed that Ag formulated with CpG induced potent B . pertussis-specific IFN-γ production and also induced IL-17A , but low concentrations of IL-4 and IL-13 ( Figure 6B ) . In contrast , immunization with Ag and alum generated B . pertussis T cells that secreted IL-4 and IL-13 , as well as IL-17A but little IFN-γ . The strongest IgG response was induced with alum as the adjuvant and these antibodies were almost exclusively IgG1 ( Figure 6C ) . Surprisingly , immunization with Ag alone did induce significant IgG1 antibody , but weak T cell responses . In contrast , CpG induced modest IgG1 , but significantly higher IgG2a and IgG2c titres than that detected with alum as the adjuvant . These findings demonstrate that switching the adjuvant from alum to CpG promotes the induction of Th1 , Th17 and IgG2 rather than Th2 , Th17 and IgG1 responses . In order to examine the relative contribution of IL-17A and IFN-γ in protection induced by Ag administered with CpG , we performed immunization and challenge experiments in WT , IL-17A−/− and IFN-γ−/− mice . The results revealed that WT and IL-17A−/− mice immunized with Ag and CpG effectively cleared the bacteria after B . pertussis respiratory challenge ( Figure 6D ) . In contrast , clearance was significantly compromised in IFN-γ−/− mice , with 100–1000 fold more bacteria in immunized IFN-γ−/− when compared with WT or IL-17A−/− mice ( Figure 6D ) , demonstrating the key role of IFN-γ but not IL-17A in immunity induced with CpG as the adjuvant . Our findings point to a key role for Th1 and Th17 cells in immunity induced by pertussis vaccines formulated with CpG and alum respectively . We and others have reported that IFN-γ-producing cells play a critical role in host immunity to B . pertussis during primary infection in part by activating macrophages to kill intracellular bacteria [10] , [11] , [36] . Here we found that mice immunized with Ag and CpG had significantly more macrophages in their lungs than non-immunized mice 3 days post B . pertussis respiratory challenge ( Figure S8 ) . This enhanced macrophage recruitment was lost in Ag and CpG-immunized IFN-γ−/− but not in IL-17A−/− mice ( Figure S8 ) , providing indirect evidence that immunization with pertussis antigens in combination with a Th1-promoting adjuvant promote recruitment of macrophages to the lungs post challenge with B . pertussis . Collectively our findings demonstrate that Th1 cells play a more critical role than Th17 or Th2 cells in host immunity to B . pertussis and have significant implication for the rational design of more effective Pa .
The significant new finding of this study is that Th17 cells mediate protective immunity induced with current alum-adjuvanted Pa . In contrast , immunity induced by infection or immunization with Pw is mediated largely by Th1 cells , with a smaller contribution from Th17 cells . Although Th2 cells are strongly induced by Pa in mice and humans , and are considered to be important in promoting antibody responses to extracellular pathogens , our study demonstrated that they are not necessary for protective immunity in a mouse model . Using a mouse respiratory challenge model , we demonstrate that transfer of Th1 or Th17 cells prior to infection of naive mice reduced the bacterial burden post challenge . We had previously reported that mice defective in IFN-γR develop disseminating lethal infection following primary challenge with B . pertussis [10] , and here we show that the bacterial burden and clearance is also significantly compromised in IL-17A−/− mice . Collectively , these studies suggest that both Th1 and Th17 cells function in natural immunity to B . pertussis . Immunization with Pw also induced Th1 and Th17 responses but studies in IFN-γ−/− mice demonstrated a dominant role for Th1 cells in mediating vaccine induced protection . In contrast a licensed Pa induced B . pertussis-specific Th17 and Th2 cells but failed to generate Th1 cells . Challenge experiments in cytokine defective mice demonstrated that IL-4 was dispensable , whereas IL-1 and IL-17A were absolutely required for protective immunity induced with Pa . Furthermore , we demonstrate that protection induced by immunization of mice with genetically detoxified PT and FHA could be enhanced by substituting alum with an adjuvant that induces Th1 as well as Th17 cells . All animal models involving inbred mice , including the one used in this study have some limitations in terms of extrapolating of experimental findings to humans . Nevertheless , we have shown that the rate of bacterial clearance in mice immunized with Pa relative to control non-immunized mice , correlates with vaccine efficacy in children [12] . Furthermore , there are significant parallels in the T cell responses in mice and humans induced by immunization or infection with B . pertussis and these can be summarized as follows: 1 ) Infection with B . pertussis induces Th1 , but not Th2 , responses in mice [10] , [27] and in humans [37]–[40] and also induces Th17 responses in both species [16] , [17] , [41] [and current study] , 2 ) Immunization with Pw induces Th1 but not Th2 responses in mice [12] , [13] , [28] and humans [8] , [9] , [42] and 3 ) immunization with Pa promotes or enhances Th2 responses in mice [12] , [13] , [28] and humans [8] , [9] , [42] , [43] . Pa also promote Th17 responses in mice [current study] , but this has not yet been fully evaluated in humans . The only slight discrepancy between mouse and human studies is that the responses induced by vaccination with Pa are mixed Th1/Th2 in human and more Th2-dominated in mice , with some reports that Pa induce Th1 responses in humans [44] . However , it has been suggested that the Th1 responses detected in certain Pa-immunized humans may result from natural acquisition following exposure to live B . pertussis [37] , [39] . Furthermore , booster immunization of 4–6 year old children primed with Pa as infants enhanced Th2 but not Th1 responses [42] , [45] , suggesting that Pa do promote similar responses in both mice and humans . Prior to the present study , the consensus view on the mechanism of protection against B . pertussis was that like vaccines against other extracellular bacteria , antibodies and Th2 responses played a central role in protective immunity generated by Pa in mice and in humans . Repeated booster immunization with Pa induces very strong B . pertussis-specific Th2 [42] , [45] and strong but transient serum IgG antibodies in children [46] and mice [14] and correlative studies showed an association between serum antibody response to PT , pertactin and fimbrae 2/3 and protection in children [7] . However , the present study shows that Th2 responses are dispensable for immunity induced by Pa in mice and that IL-17A plays an essential role in protection induced with alum-adjuvanted Pa . Furthermore , FHA , which is considered to be the least important antigen from the perspective of antibody production [7] , was a major target for Th17 cells . It has previously been reported that Th17 cells are involved in adaptive immunity to Pseudomonas aeruginosa [47] , Staphylococcus aureus , Candida albicans [48] and Helicobacter pylori [49] . It has also been reported that Th17 cells mediate heterologous protection against Klebsiella pneumonia induced by nasal immunization with heat-killed bacteria [50] . In addition , it has been demonstrated that immunization of mice with the Mycobacterium tuberculosis ESAT-6 peptide mixed with the TLR4 agonist MPL in combination with trehalose dimycolate analog as the adjuvant/delivery system induced IL-17-producing cells which conferred protection by helping to recruit IFN-γ-secreting Th1 cells to the lungs [51] . However , to our knowledge this is the first study to identify a role for IL-17A in protective immunity against B . pertussis in mice induced with a licensed alum-adjuvanted human vaccine . The induction and expansion of Th17 cells involves a number of inflammatory cytokines , including IL-1β and here we found that alum-adjuvanted Pa or alum alone promoted innate IL-1β production . It has been reported that IL-1β , which is processed by caspase-1 , did not mediate the adjuvant activity of alum in vivo , however these conclusions were based on the premise that alum promotes immunity via the induction of Th2 and antibody responses . It has been demonstrated that mice defective in caspase-1 or Nlrp3 had reduced Th2 and antibody responses to antigens administered with alum [20] , [21] . However , it was later reported that while the induction of IL-1β by alum was dependent on Nlrp3 , the enhancement of antibody responses with alum in vivo were independent of Nlrp3 [22] , [52] . The present study demonstrated that Nlrp3 was required for the induction of IL-1β production by DC in vitro with alum or alum-containing Pa , but only played a minor role in the induction of IL-1β , IL-17 and protective immunity induced with Pa in vivo , suggesting involvement of Nlrp3-independent pathways in cells other than DCs . This is consistent with a role for Nlrp3 and caspase-1 independent IL-1β in host defences against Mycobacterium tuberculosis in vivo . [53] , [54] . Consistent with a previous report [55] , we found Pa-immunized IL-1RI−/− mice had a significantly higher bacterial burden after B . pertussis challenge than WT mice and this was associated with significantly reduced antigen-specific IL-17A production . Collectively these findings suggest that the adjuvant activity of alum in enhancing the immunogenicity of Pa is mediated at least in part by induction of innate IL-1 , which in turn drives the induction of protective Th17 cells . A significant new finding of our study is that Th17 cells mediate protective immunity induced by Pa through recruitment of macrophages and neutrophils to the lungs , which phagocytose and kill B . pertussis . Immunization with Pa was associated with rapid induction of chemokines and recruitment of neutrophils and macrophages to the lungs after B . pertussis challenge and these inflammatory responses and associated bacterial clearance were significantly reduced in IL-17A−/− mice . We had previously reported that IL-17A and IFN-γ enhance killing of B . pertussis by macrophages [15] . Here , we found that IL-17A and IFN-γ also promoted killing of B . pertussis by neutrophils in vitro . Neutrophils do not appear to play a critical role in clearing a primary infection with B . pertussis in naïve mice , but are essential for control of B . bronchiseptica and play a role in controlling B . pertussis infection in immune mice [56] , [57] . Although antibody did not significantly enhance killing of B . pertussis by neutrophils in our in vitro assay , our data do not rule out a role for antibodies , especially murine IgG2 , in protective immunity against B . pertussis . Although Pa have significantly improved safety profiles over Pw that they replaced and protect a significant percentage of children against a life threatening disease , there is an increasing incidence of pertussis in many developed countries [4] . In an attempt to limit the spread of B . pertussis , a number of countries have introduced booster vaccinations with Pa for 5–6 year olds , adolescents and even adults . However , a recent report has suggested protection from whooping cough in children that had received 5 doses of Pa is relatively short-lived and wanes substantially each year [6] . Furthermore , repeated boosting of Th2 or Th17 responses may not be desirable , since these responses can mediate hypersensitivity/allergy or autoimmunity when directed against allergens or self antigens respectively [58] . Indeed there is already evidence of hypersensitivity reactions in pre-school children following the fifth dose of Pa [59] . The corollary to this is that efficacy of a vaccine that relies heavily on IL-17A to confer protective immunity may be compromised in patients treated with IL-17A targeted drugs , which are in late stage clinical development for autoimmune diseases [60] . The induction of Th1 rather than Th2 responses by Pa delivered with an appropriate Th1-promoting adjuvant may not only be safer but more effective than alum-adjuvanted Pa . We found that Pw confer a high level of immunity by inducing Th1 cells , with a smaller contribution by Th17 cells . It has previously been shown that addition of the TLR9 agonist CpG to an alum-adjuvanted pertussis vaccine enhanced IgG1∶IgG2a antibodies , providing indirect evidence of enhanced Th1 responses [61] . In addition it has recently been reported that CpG enhances the protective efficacy of B . pertussis Ag when administered i . n . or i . p , with aluminum hydroxide [62] . However , immunization with B . pertussis Ag and CpG ( 30 µg/dose; without alum ) did not enhance IFN-γ production or protective efficacy over that observed with Ag formulated with alum [62] . This is surprising given the previous reports that CpG enhances Th1 response to soluble antigens [35] . Here we found that alum-adjuvanted Pa induced weak Th1 responses , but formulation of an experimental Pa with CpG ( 50 µg/dose; without alum ) promoted Th1 as well as Th17 responses and conferred a significantly greater level of protection against B . pertussis . Furthermore , protection induced with the CpG-adjuvanted experimental vaccine was significantly compromised in IFN-γ−/− , but retained IL-17A−/− mice . These findings suggest that pertussis vaccine formulations that employ adjuvants that promote Th1 responses , such as TLR agonists , should be evaluated as a safe and more effective alternative to current alum-adjuvanted Pa for use in humans .
All mice were maintained according to EU regulations and experiments were performed under licence from the Irish Department of Health and Children and with approval from the Trinity College Dublin Bioresources Ethics Committee . IL-1RI−/− , Nlrp3−/− , IL-17A−/− , IL-4−/− , IFN-γ−/− and C57BL/6 WT mice were bred in house from established colonies and housed under specific pathogen free conditions . IL-17A−/− mice were provided by Yoichiro Iwakura , Centre for Experimental Medicine and Systems Biology , Institute of Medical Science , University of Tokyo , Japan . The Pa used in this study was a commercially available INFANRIX ( GSK; diphtheria and tetanus toxoids and acellular pertussis adsorbed to alum; the pertussis component comprising detoxified PT , FHA and pertactin ) . The Pw used in this study ( the third international standard preparation , 88/522 from NIBSC , Herts . , UK ) was a thiomersal killed B . pertussis vaccine . Mice were immunized i . p . or i . m . ( quadriceps muscle ) twice ( wk 0 and 4 ) with 0 . 2 human dose of Pa or Pw and were challenged with B . pertussis by aerosol inoculation or sacrificed 2 wks after second immunization . In experiments designed to examine the effect of CpG versus aluminium hydroxide ( alum ) as the adjuvant we use an experimental laboratory prepared Pa using two purified antigens , detoxified PT and FHA present in all 2 and 3-component Pa ( 1 and 2 . 5 µg/mouse respectively ) . The PT was GMP-grade genetically detoxified PT ( PT-9K/129G ) , supplied by Novartis Vaccine , Siena , Italy . FHA was purchased from Kaketsuken , Kumamoto , Japan . Both preparation were highly purified , as determined by SDS gel chromatography and were free of detectable LPS . Phosphorothioate-stabilized oligodeoxynucleotide-containing CpG motifs ( CpG ) ( 5′tccatgacgttcctgatgc ) was obtained from Sigma-Genosys Ltd , Cambridge , UK and used at 50 µg/mouse dose . Aluminum hydroxide ( alhydrogel; referred to as alum ) was obtained from Brenntag Biosector , Friederikssund , Denmark and used at 100 µg/mouse dose . Respiratory infection of mice was performed by aerosol challenge as previously described [63] . The course of B . pertussis infection was followed by performing CFU counts on lungs from groups of 4–5 mice at intervals after challenge . The lungs were aseptically removed and homogenised in 1 ml of sterile physiological saline with 1% casein on ice . Undiluted and serially diluted homogenate ( 100 µl ) from individual lungs was spotted in triplicate onto Bordet-Gengou agar plates , and the number of CFU was calculated after 5 days incubation at 37°C . The limit of detection was approximately 0 . 3 log10 CFU per lung for groups of 4 mice at each time point ( indicated by a dotted line on each CFU curve ) . Mononuclear cells were prepared from the lungs of naive and B . pertussis infected mice by mechanical disruption of lung tissue [63] . Lung mononuclear cells or spleen cells ( 1–2×106/ml ) were cultured at 37°C and 5% CO2 with heat killed B . pertussis or purified FHA . Stimulation with PMA ( 250 ng/ml; Sigma ) and anti-mouse CD3 ( 1 µg/ml; Pharmingen , San Diego , USA ) or medium only was used as positive and negative controls respectively . Supernatants were removed after 72 h and IL-4 , IL-13 , IL-17 and IFN-γ concentrations determined by two-site ELISA . To determine cells infiltrating the lung following infection isolated lung mononuclear cells were isolated as above , washed and blocked with Fcγ block ( 1 µg/ml; BD Pharmingen ) before surface staining with CD11b , GR1 and F4/80 . Neutrophil numbers were determined by gating on GR1+ and CD11b+ while macrophages numbers were determined by gating on F4/80+ CD11b+ cells . For intracellular cytokine staining , isolated cervical lymph nodes ( 2×106 cells/ml ) were stimulated for 5 h with PMA , ionomycin in the presence of brefeldin A ( 5 µg/ml ) . Alternatively lung mononuclear cells were incubated for 1 h with brefeldin A ( 5 µg/ml ) only . Cells were washed and blocked with Fcγ block ( 1 µg/ml; BD Pharmingen ) before extracellular staining for surface CD3 and CD4 ( BD Pharmingen ) . Cells were then fixed and permeabilized ( Fix and Perm cell permeabilization kit; Caltag Laboratories ) and stained for intracellular IL-17A and IFN-γ . Flow cytometric analysis was performed using a CyANADP Flow Cytometer ( DakoCytomation ) and analysed with FlowJo software , with gating set on fluorescence minus one controls . Serum antibody responses to B . pertussis were quantified by ELISA using plate-bound heat-killed B . pertussis or FHA ( 5 µg/ml ) . Bound antibodies were detected using biotin-conjugated anti-mouse IgG , IgG1 , IgG2a or IgG2c antibodies ( Caltag ) and peroxidase-conjugated streptavidin ( BD Pharmingen ) . Antibody levels are expressed as the mean endpoint titre ( ± SE ) , determined by extrapolation of the linear part of the titration curve to 2 SE above the background value obtained with non-immune mouse serum . Bone marrow-derived DC were prepared by culture with GM-CSF as previously described [64] . DC were cultured with alum ( 125 µg/ml ) , Pa ( 0 . 025 , 0 . 1 and 0 . 4 IU/ml ) and ATP ( 5 mM; Sigma ) , or medium only , with or without the caspase-1 inhibitor Ac-YVAD-cmk ( 40 µM; Calbiochem ) . Supernatants were recovered and IL-1β concentrations determined by ELISA ( R&D Systems ) . Neutrophils were collected from the peritoneal cavity of WT mice 18 hours following i . p . injection of 500 µl of 9% casein ( Sigma ) and were purified by centrifugation over 62% percoll ( GE Healthcare ) yielding a 97% pure population . B . pertussis ( 106/test ) were incubated with 10% normal mouse or immune mouse serum for 20 minutes at 37°C after which neutrophils ( 106/test ) together with 50 ng/ml of recombinant IL-17A , IL-17F ( eBioscience ) or IFN-γ ( R&D systems ) were added and incubated with shaking . After the appropriate times ice cold dH2O was used to lyse the cells and a CFU count determined as described above . One-way analysis of variance ( ANOVA ) was used to test for statistical significance of differences between more than two experimental groups . The student's t test was used for analysis when two groups were compared . | The bacterium Bordetella pertussis causes whooping cough , a severe and often lethal respiratory infection in humans . The disease was largely controlled through vaccination with whole cell pertussis vaccines ( Pw ) . However , Pw had side effects and were replaced in developed countries in the 1990s with safer acellular pertussis vaccines ( Pa ) . Unfortunately this has now been linked with a recent resurgence of whooping cough . We have used a mouse model to examine the mechanism of host immunity against B . pertussis . We examined the type of immune responses induced with Pa compared with Pw in an attempt to identify its shortcomings and to design a more effective vaccine . Traditionally it had been considered that antibodies mediate protection induced with pertussis vaccines . However , we found that blood lymphocytes , in particular a subpopulation of T cells called Th17 cells that secrete a cytokine called IL-17 , play a critical role in host immunity induced by Pa . In contrast , Pw induce Th17 cells but also another T cell subtype called Th1 cells , which are also required for optimum immunity . Finally , we rationally designed a new vaccine using a formulation that induces Th1 cells and found that this was highly effective in conferring protective immunity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biology"
] | 2013 | Relative Contribution of Th1 and Th17 Cells in Adaptive Immunity to Bordetella pertussis: Towards the Rational Design of an Improved Acellular Pertussis Vaccine |
An increasing number of neuroscience papers capitalize on the assumption published in this journal that visual speech would be typically 150 ms ahead of auditory speech . It happens that the estimation of audiovisual asynchrony in the reference paper is valid only in very specific cases , for isolated consonant-vowel syllables or at the beginning of a speech utterance , in what we call “preparatory gestures” . However , when syllables are chained in sequences , as they are typically in most parts of a natural speech utterance , asynchrony should be defined in a different way . This is what we call “comodulatory gestures” providing auditory and visual events more or less in synchrony . We provide audiovisual data on sequences of plosive-vowel syllables ( pa , ta , ka , ba , da , ga , ma , na ) showing that audiovisual synchrony is actually rather precise , varying between 20 ms audio lead and 70 ms audio lag . We show how more complex speech material should result in a range typically varying between 40 ms audio lead and 200 ms audio lag , and we discuss how this natural coordination is reflected in the so-called temporal integration window for audiovisual speech perception . Finally we present a toy model of auditory and audiovisual predictive coding , showing that visual lead is actually not necessary for visual prediction .
Sensory processing has long been conceived as modular and hierarchic , beginning by monosensory cue extraction in the primary sensory cortices before higher level multisensory interactions took place in associative areas , preparing the route for final decision and adequate behavioral answer . However , it is now firmly established that low-level multisensory interactions are much more pervasive than classical views assumed they were and affect brain regions and neural responses traditionally considered as modality specific [1] , [2] . Restricting to audiovisual interactions in speech perception , direct connections have been displayed between primary auditory cortex and primary visual cortex ( e . g . [3] on macaques ) , and electrophysiological data on speech perception display early influence of the visual component of speech stimuli on auditory evoked response potentials ( ERPs ) . Indeed , there appears a decrease in amplitude and latency of the first negative peak N1 and the second positive peak P2 , 100 to 200 ms after the acoustic onset , when the visual component is present [4] , [5] . It is still under debate to determine the specific role of direct connections between primary sensory cortices vs . the role of associative cortex and particularly the superior temporal sulcus in these early interactions [6]–[8] . The computational nature of audiovisual interactions is now the focus of a large number of recent papers . Capitalizing on the natural rhythmicity of the auditory speech input , it has been suggested [9] , [10] that the visual input could enhance neuronal oscillations thanks to a phase-resetting mechanism across sensory modalities . This has led to various experimental demonstrations that visual speech improves the tracking of audiovisual speech information in the auditory cortex by phase coupling of auditory and visual cortices [11] , [12] . A number of these studies have proposed predictive coding as a possible unifying framework for dealing with audiovisual interactions . Predictive coding posits that neural processing exploits a differential coding between predicted and incoming signals , with decreased activity when a signal is correctly predicted [13] , [14] . Visual prediction would be responsible for early modifications in auditory ERPs evoked by visual speech decreasing latency and amplitude of N1 and P2 ( e . g . [5] , [7] ) . This has led to recent proposals about the role of specific components in neural oscillations respectively conveying top-down predictions and bottom-up prediction errors in audiovisual speech processing [8] , [15] . The previously mentioned studies capitalize on the underlying audiovisual structure of speech stimuli , that is the way sounds and sights provided by the speaker are comodulated in time ( so that their phase can indeed be coupled ) and more generally how one modality provides adequate information for partial prediction of the other modality . It is actually known since long that the auditory and video streams are related by a high level of cross-predictability related to their common underlying motor cause . This is displayed in a number of studies about audio-visual correlations between various kinds of video ( e . g . lip parameters , facial flesh points , video features extracted from the face ) and audio ( acoustic envelope , band-pass filter outputs , spectral features ) parameters [16]–[20] . In a recent and influential paper published in this journal , Chandrasekaran et al . [21] present a number of analyses about the “natural statistics of audiovisual speech” , based on various databases in different languages ( British and American English , and French ) , with four major results: firstly , there is a robust correlation in time between variations of mouth opening and variations of the acoustic envelope; secondly , focusing the acoustic envelope to narrow regions in the acoustic spectrum , correlation is maximum in two regions , one around 300–800 Hz , typically where is situated the first vocal tract resonance ( formant ) F1 , and the other around 3000 Hz interpreted by the authors as corresponding to the second and third resonances F2 and F3; thirdly , temporal comodulations of the mouth and acoustic envelope appear in the 2–7 Hz frequency range , typically corresponding to the syllabic rhythm; last but not least in the context of the present paper , “the timing of mouth movements relative to the onset of the voice is consistently between 100 and 300 ms” ( penultimate sentence of the paper abstract ) . Since the publication of this paper and systematically referring to it , an increasing number of neuroscience papers – including some of those cited previously – capitalize on the assumption that visual speech would be typically 150 ms ahead of auditory speech . Let us mention a few quotations from these papers: “In most ecological settings , auditory input lags visual input , i . e . , mouth movements and speech associated gestures , by ∼150 ms” [7] , [8]; “there is a typical visual to auditory lag of 150–200 ms in face-to-face communication” [22]; “articulatory facial movements are also correlated with the speech envelope and precede it by ∼150 ms” [12] . The invoked natural audiovisual asynchrony is used in these papers in support to development on models and experiments assessing the predictive coding theory . The assumption that image leads sound plays two different roles in the above mentioned neuroscience papers . It is sometimes used as a trick to demonstrate that the visual stimulus plays a role in modulating the neural auditory response , rightly capitalizing on a situation where a consonant-vowel ( CV ) sequence ( e . g . “pa” or “ta” ) is produced after a pause . In this case , the preparatory movement of the mouth and lips is visible before any sound is produced , hence visual prediction can occur ahead of sound and results in visual modulation of auditory ERPs [4] , [5] , [7] . The second role is more problematic . Considering that there would be a systematic and more or less stable advance of vision on audition around 150 ms , it is proposed that this situation would play a role in the ability to use the visual input to predict the auditory one all along the time . Audiovisual asynchrony is implicitly incorporated in a number of models and proposals . However , as we will see in the next section , the situation studied in [21] is very specific , characteristic of a CV sequence produced in isolation or at the beginning of an utterance after a pause . The objective of the present paper is to show that , while the method proposed by Chandrasekaran et al . to estimate audiovisual delays is adequate for the onset in preparatory sequences or the start of a speech utterance , in chained sequences which actually provide the most general case in speech communication , the method should be modified . Furthermore , if an appropriate method is used , delays actually vary in a different range from the one they propose – with the consequence that “there is no 150 ms lead of visual speech on auditory speech” . The rationale in the measure of asynchrony proposed by Chandrasekaran et al . is based on the notion of preparatory gestures ( Figure 1 ) . This is also the case of the N1-P2 studies mentioned previously ( e . g . [5] , [8] ) . This can be related to a rather classical analogy , namely the movement of a hammer towards a table ( Figure 1a ) . To produce a sound with a hammer , one must previously realize a downward stroke and the onset of this downward stroke is visible much before the hammer touches the table and makes a sound . Notice that in this scene , one could define actually two visible events , one at the onset of the downward stroke and one at the instant when the hammer touches the table; and only one auditory event , the sound onset , which is actually perfectly synchronous with the second visual event . The downward stroke may be called a “preparatory gesture” in that it prepares the sound and hence predicts something about it ( its time of arrival , and also its acoustic content since a subject looking at the hammer going towards the table knows the kind of sound which will be produced soon ) . It is exactly the same for preparatory lip gestures before “p” at the beginning of a speech utterance ( Figure 1b ) : when the lips begin to close , a subject looking at the speaker knows that they will soon join together for a lip closure , and she/he can predict rather accurately when will sound occur and what will be its spectrum ( the typical flat low-frequency spectrum of a bilabial burst [23] ) . Here again , there are two visual events , namely the onset of the lip closing gesture and the further onset of the lip opening gesture , and only one auditory event , the burst onset , quite synchronous with the second visual event . Notice that the analogy between the preparatory gestures for the hammer and for speech is not perfect . Indeed , the sound is produced by the hammer at the end of the downward stroke , while for speech the lips must open again . There is actually a complex coordination between larynx , lungs and lips to achieve the adequate aerodynamic strategy [24] , which fixes rules about the duration of lip closure before lip opening . But the audiovisual asynchrony involved in preparatory gestures for both hammer and speech are similar: in both cases , audiovisual asynchrony is assessed by the duration between two different events , the onset of the preparatory gesture for the visual channel and its offset for the auditory channel . Therefore it appears that the crucial aspect of preparatory gestures is that they are visible but produce no sound . This could be different , actually . Consider for example what happens if you replace the hammer by a whip or a flexible stick . Now the downward stroke produces a whistling sound ( which also predicts the sound produced when the whip or stick touches the table ) . There are now two auditory events , just as there are two visual events , and for both pairs of audiovisual events ( at the beginning and end of the visual stroke ) the auditory and visual events are quite in synchrony . This leads us towards another kind of gestures that we propose to call “comodulatory gestures” since these gestures produce both auditory and visual stimuli more or less in synchrony all along the time ( Figure 2 ) . Comodulatory gestures are actually by far the most common gestures in speech . Here we should move towards another analogy that is a balloon in which one adjusts the mouthpiece . When its size increases or decreases , shape , volume and pressure change leading to more or less synchronous auditory and visual events for both opening and closing phases ( Figure 2a ) , just as opening and closing the lips while vocalizing produces auditory and visible events quite in synchrony ( Figure 2b ) . In the remaining of this paper we present simple audiovisual data on plosive-vowel syllables ( pa , ta , ka , ba , da , ga , ma , na ) , produced either in isolation or in sequence . We show that when syllables are produced in isolation , preparatory gestures provide audiovisual asynchronies quite in line with those measured in [21] . However , when syllables are chained in sequences , they provide comodulatory gestures in which audiovisual synchrony is actually precise , contrary to the data provided on similar sequences in [21] , just because the measure of audiovisual asynchrony is different . In such cases , there are actually auditory events that were not taken into account in the original paper , and these need to be taken into account if one is talking about asynchrony . After presenting Methodology and Results , we discuss how natural coordination between sound and image can actually produce both cases of lead and lag of the visual input . We relate the range of leads and lags to the so-called temporal integration window for audiovisual speech perception [25] . We propose that the “visual lead” hypothesis , wrong in many cases , is actually not necessary to deal with audiovisual predictability , and we illustrate this by briefly introducing a simple audiovisual prediction model dealing with the speech sequences studied previously . We conclude by some methodological and theoretical remarks on neurophysiological developments about audiovisual predictability in the human brain .
In the experimental work we focus on audiovisual temporal relationships in CV sequences where C is a voiced , unvoiced or nasal stop consonant that is , for English or French ( the two languages considered in [21] ) , one of the sounds /p t k b d g m n/ , and V is the open vowel /a/ . We consider both CV sequences produced in isolation and chained sequences VCVCVCV . This corpus is very simple though sufficient to illustrate the difference between preparatory gestures – for isolated syllables – and comodulatory gestures – for chained syllables . The /a/ context in which the plosives /p t k b d g m n/ are produced is selected because it provides a large gesture amplitude providing more salient trajectories both in the visual and auditory modality . We will consider more general phonetic material in the discussion . We recorded a small database of 6 repetitions of 8 syllables /pa ta ka ba da ga ma na/ uttered by a French speaker either in isolation /Ca/ or in sequence /aCa/ . The syllables were produced in a fixed order at a relatively slow rhythm ( around 800 ms per syllable ) . In the “isolated syllables” condition , syllables were embedded in silence: /pa#ta#ka#ba#da#ga#ma#na/ where /#/ means a silence ( typically 500 ms silence between two consecutive syllables ) . In the “chained syllables” condition , they were produced in the same order though with no silence between syllables: /apatakabadagamana/ . The recording was done with a PAL camera at 50 Hz . The recording set up was based on the classical paradigm we use in Grenoble since years [26] , [27] with blue make up applied on the lips . For each image , we extracted automatically and precisely the lip contours by applying a Chroma Key process extracting blue areas on the face . The lips parameters were extracted every 20 ms , synchronously with the acoustic signal , which is sampled at 22 . 05 kHz .
We display on Figure 5 the data for isolated syllables . In this case , where there is no audible event for closure , we report the same measure as in [21] that is the delay between the first visible event , CVL , and the first audible event , OAI or OAF . There is a very large anticipation , which actually reaches values much larger than 150 ms here ( and which may reach 400 ms in some cases ) . These values are compatible with the range 100-300 ms proposed in [21] , the more so considering that the measure used by the authors for detecting visual events ( half open point in the lip closing trajectory , while we used the onset of the closing phase ) would produce values lower than the ones in Figure 5 . We display on Figure 6 typical audiovisual sequences for all types of chained syllables ( with a zoom around the consonant ) . It clearly shows that there is comodulation of the auditory and visual information , with audible and visible events for both closing and opening phases . The event detection is sometimes not straightforward or not very precise in time ( e . g . detection of CAI for /ata/ or /ada/ ) , which is quite classical in this type of stimuli , and gross trends are more important that precise values in the following . We display on Figure 7 the data about temporal coordination between audio and visual events for either closing ( Figure 7a ) or opening ( Figure 7b ) in the case of chained sequences . The mean delay between visual and acoustic events at the closure ( in the /aC/ portion , Figure 7a ) varies between −20 ms and −40 ms for intensity ( CVL-CAI , in green ) and reaches values from −40 to −80 ms for formants ( CVL-CAF , in red ) . This means that there is a small lead of the visual channel compared to the audio channel ( where information is available on intensity before formants ) . But this lead is much smaller than the 150 ms lead mentioned in [21] , and there are actually cases where audio and video information are available more or less in synchrony , e . g . for /ad/ , /ag/ or /ak/ where the tongue gesture towards the voiced plosive decreases intensity or formants while jaw may stay rather stable , and hence lip area does not decrease much – which prevents early video detection . In the opening phase ( Figure 7b ) the synchrony is even larger . Concentrating on the delay between labial and intensity events ( OVL-OAI , in green ) we actually observe an almost perfect synchrony for labials ( /p b m/ ) . This is trivial: as soon as the lips begin to open , the sound drastically changes , from silence ( for /p/ ) or prevoicing ( for /b/ ) or nasal murmur ( for /m/ ) to the plosive burst . For velars /k g/ there is actually a clear lead of the audio channel , since the first tongue movement producing the plosive release is done with no jaw movement at all and hence before any labial event is actually detectable: the audio lead may reach more than 20 ms ( see examples in Figure 6 ) . Notice that while the video sampling frequency at 50 Hz can make the detection of the opening event for bilabials a bit imprecise with a precision around 10 ms for very quick gestures , the variations of lip area for dentals or velars is smooth and hence imprecision in event detection cannot explain such an audio lead . Therefore the discrepancy with [21] is clear for chained syllables , just because this corresponds to what we called comodulatory gestures , for which we argue that a different measure of the audiovisual asynchrony should be used .
The experimental results presented previously show that for isolated syllables associated with preparatory gestures , our measure of audiovisual asynchrony provides quantitative estimates from 200 ms to 400 ms of visual lead ( Figure 5 ) . This is in line with the 100 to 300 ms visual lead proposed in [21] , the more so considering that the estimate of the visible onset for lip closure in [21] is done at the mid closing phase – while we prefer detecting the first visible event that is at the very beginning of the lip closure phase , typically 100 ms before . The coherence of both sets of measures was expected considering that the same definition of asynchrony for preparatory gestures is used in both papers , between the first visible event ( onset of lip closing phase ) and the first auditory event ( plosive burst at labial release ) . However the data are quite different for chained sequences associated with comodulatory gestures . In this case the range of asynchronies is much more restricted and more centered around 0 , from 70 ms visual lead to 20 ms audio lead when auditory events are detected on intensity , auditory events detected on the formant trajectory being somewhat delayed in respect to intensity ( Figure 7 ) . Mean video lead amounts to 35 ms in the closing phase and 0 ms in the opening phase for intensity , 60 ms in the closing phase and less than 10 ms in the opening phase for formants . Therefore the departure between our data and those proposed in [21] is now important . This is not due to variability in the speech material , but to a difference in the measure proposed for assessing audiovisual asynchrony . As explained in Figure 4 , the measures differ hence their results also differ . Speech gestures in chained sequences typically produce both auditory and visual events all along the time ( see Figure 6 ) hence resulting in a rather precise audiovisual synchrony in most cases . Preparatory gestures do exist in speech communication , and ERP studies rightly capitalized on this experimental situation in which the gap between the first visible and the first auditory event may be quite large and able to lead to significant influence of the visual input on the electrophysiological response in the auditory cortex , for both speech [5] , [8] and non-speech stimuli [29] , [30] . Notice that this may actually depend on the prephonatory configuration: if somebody keeps the lips closed while listening to the interlocutor , there will actually be no preparatory gesture before an initial bilabial sound such as /b/ or /m/ , and hence there will be no visual lead at all in this case . One could even imagine a reverse situation in which a speaker keeps the lips closed and systematically signals her/his turn taking by a backchannel signal “hmm” ( which is not rare ) : in this case the preparatory gesture would be actually audible and not visible , leading to an auditory lead in the preparatory phase . However , most of the speech material is made of comodulatory gestures . Of course , speech utterances involve a range of phonetic configurations much larger than the /Ca/ sequences that were studied in this paper . This variety of configurations leads to a variety of situations in terms of audiovisual asynchronies . This is where the analogy we proposed previously with the deflating balloon being both audible and visible reaches some limits: actually , not every action realized on the vocal tract is always either audible or visible , which may lead to delays between perceivable auditory or visible cues for a given speech gesture . A first general property of speech concerns anticipatory coarticulation – much more relevant and general than preparatory movements discussed in [21] . This relates to articulatory gestures towards a given phonetic target , which can begin within a previous phoneme . Anticipatory coarticulation generally capitalizes on a property of the articulatory-to-acoustic transform , in which an articulatory gesture has sometimes no or weak effect on the sound and hence can be prepared in advance without audible consequences . A typical example concerns the rounding gesture from /i/ to /y/ or /u/ in sequences such as /iC1C2…Cny/ or /iC1C2…Cnu/ with a variable number of consonants C1…Cn not involving a specific labial control ( e . g . /s t k r/ ) between the unrounded /i/ and the rounded /y/ or /u/ . In this case the rounding gesture from /i/ towards /y/ or /u/ can begin within the sequence of consonants /C1C2…Cn/ , and hence anticipate the vowel by 100 to 300 ms [31] . Various sets of data and various theoretical models of this anticipatory coarticulation process have been proposed in the literature [32]–[36] . In such cases the rounding gesture can hence be visible well before it is audible . So there are cases where visible information is available before auditory information ( e . g . in /iC1…Cnu/ sequences ) , others where vision and audition are quite synchronous ( e . g . in /aCa/ sequences ) , and there are also cases where audition may actually lead vision as was shown e . g . in Figure 7 . But the next question is to know if the auditory and visual systems are able to process the information efficiently as soon as it is available . This is actually not always the case , and in gating experiments on the visual vs . auditory identification of coarticulated sequences , Troille et al . [37] display in some configurations a lead of audition on vision which can reach up to 40 ms , because of the poor visibility of some articulatory gestures . This leads the authors to claim that they have discovered a case where “speech can be heard before it is seen” . In summary , there are actually a variety of situations from audio lead ( estimated to 40 ms in [37] ) to visual lead ( which can reach more than 200 ms ) . In their study of mutual information between audio and video parameters on speech sequences , Feldhoffer et al . [38] show that mutual information is maximal for some audio and video parameters when it incorporates a video lead up to 100 ms . In audiovisual speech recognition experiments , Czap [39] obtains a smaller value , recognition scores being higher with a small global video lead ( 20 ms ) . Altogether , these global estimations are concordant with the classical view that “in average , the visual stream may lead the auditory stream” , which is generally advocated by specialists of audiovisual speech perception ( e . g . [40] , [41] ) . However , the “average” view hides a large range of variations , typically inside a window between 40 ms audio lead to 200 ms visual lead in the phonetic content of normal speech communication . A large number of recent studies have attempted to characterize the temporal integration window in various kinds of multisensory interactions . This typically involves two kinds of paradigms . Firstly , evaluation of intersensory synchrony may be based on either simultaneity or temporal order judgment tasks ( see a recent review in [42] ) . Secondly , the “multisensory temporal binding window” describes the range of asynchronies between two modalities in which a fused percept may emerge [43] . The “audiovisual temporal integration window” is well described for speech perception ( e . g . [44] , [45] ) . Van Wassenhove et al . [25] compared estimates of audiovisual temporal integration window based on either simultaneity perceptual judgments or regions where the McGurk effect seems to stay at a maximal value . They show that these various estimates converge on an asymmetric window between about 30 ms audio lead and 170 ms audio lag . This provides a set of values rather coherent with the range of possible asynchronies in the speech material itself . Small audio leads may occur because of the lack of visibility of certain audible gestures , as shown in Figure 7 or in gating experiments [37] . Large video leads are mostly due to labial anticipatory coarticulation and described in many studies [31]–[36] . A tentative interpretation is that the perceptual system has internalized this range through a learning process . This is in line with the so-called “unity assumption” [46] according to which subjects would naturally bind together multisensory stimuli referring to a common cause , which would lead to both fused percepts and decreased ability to detect temporal asynchronies [47] . We speculate that unity assumption is based on a statistical learning of the comodulation properties of the auditory and visual streams in the speech natural environment , naturally providing an asymmetrical window around the range [−30 ms , +170 ms] . The asymmetry of the temporal integration window has been the topic of much discussion – including assumptions about the difference between optic and acoustic wave speeds , which cannot however explain such a large asymmetry: a speaker 10 m apart from a listener would not provide more than 30 ms visual advance ! We argue here that the psychophysical asymmetry just mirrors the natural phonetic asymmetry , according to which there are plenty of cases of large visual anticipation due to coarticulation – typically in the 100 to 200 ms range – and less cases of auditory anticipation , in a smaller range – typically less than 40 ms as displayed in our data in Figure 7 or in gating data [47] . But , once again , this does not mean that there is a constant visual lead , but rather a range of audiovisual asynchronies mirrored in the temporal integration window . Recent data on the development of the audiovisual temporal integration window fit rather well with this proposal . Indeed , these data show that the window is initially quite large and then progressively refined by “perceptual narrowing” in the first months of life [48] . The window actually appears to stay rather wide and symmetrical until at least 11 years of age [49] . It is only after this age that the left part of the window ( for auditory lead ) refines from 200 ms to 100 ms , which is proposed by the authors as the typical value for adults ( the fact that these values are larger than in [25] likely comes from the use of a different criterion to define binding windows from simultaneity curves ) . On the contrary , the right part of the window stays stable . The interpretation is that the large initial symmetric window [−200 ms , +200 ms] is progressively tuned to the window characteristic of the speech input , asymmetric in nature . The fact that learning the asymmetrical pattern occurs so late may appear surprising , but it is in fact compatible with data showing that the maturation of the McGurk effect is not complete before at least 8 years of age for native stimuli and even later for non-native stimuli [50] . There is also a rather large deal of variations of audiovisual temporal integration window from one subject to another [43] . These variations respect the asymmetry trend , though with large variations in quantitative values . The fact that these variations are correlated with the results of various fusion paradigms suggests that inter-individual differences could be related with specific weights attributed by subjects to one or the other modality [51] , [52] . Interestingly , it also appears a large ability to tune and decrease the integration window with auditory or visual experience [53] , [54] , including the possibility to decrease the asymmetry and specifically decrease the large visual-lead part of the window , which suggests that the integration window actually combines stimulus-driven content with individually-tuned perceptual experience . The data recalled in the previous section rule out over-simplistic claims about audiovisual predictability . Does it raise a problem for predictability in general ? The answer is clearly no . The reason is that predictability does not require asynchrony . Actually , a pure auditory trajectory may provide predictions on its future stages , and the visual input may enhance these predictions , since it is naturally in advance on future auditory events , though not systematically in advance on present ones . This is illustrated on the toy model presented in [55] and sketchily introduced here under ( see a detailed presentation in the Supplementary Text S1 ) . The model was developed for dealing with a corpus of repetitions of sequences /aba/ , /ada/ and /aga/ uttered by a male French speaker . A predictive coding model was developed to provide guesses about the closure point of the acoustic trajectory /aC/ ( with C one of the plosives /b , d , g/ ) from a given point of the trajectory . We implemented such a model within a Bayesian probabilistic framework , comparing predictions provided by audio-alone inputs with predictions provided by audiovisual inputs . Importantly , audiovisual inputs were shown to produce better predictions , providing values closer to the actual endpoint than with audio-only inputs . This shows that the visual component provides information able to improve predictions . This toy model is of course highly oversimplified in respect to what should be a reliable system dealing with the whole complexity of speech . However it presents the interest to show that the visual input may strongly improve predictions , in spite of the close synchrony of basic temporal events in the auditory and visual streams , according to the data presented in the Results section . In a word , there is no theoretical requirement for visual lead to argue that visual predictive coding could be at work in the sensory processing of speech in the human brain . The impressive advances of neurosciences on the processing of speech in the human brain , sometimes simplify the complexity of speech , and miss or forget a number of evidence and facts known from long by phoneticians – on the structure of phonetic information , on the auditory and visual cues , on some major principles of speech perception and production . In consequence , there is a serious risk that these advances oversimplify “much of the known complexity of speech as [it] is spoken and of speakers as they speak” [56] . This paper attempts to make clear that the view that vision leads audition is globally oversimplified and often wrong . It should be replaced by the acknowledgement that the temporal relationship between auditory and visual cues is complex , including a range of configurations more or less reflected by the temporal integration window from 30 to 50 ms auditory lead to 170 to 200 ms visual lead . It is important to recall that fortunately , this caveat does not put in question the experimental studies that capitalized on the presumed “150-ms video lead” to assess audiovisual interactions in EEG or MEG data . Indeed , all these studies ( e . g . [4] , [5] , [7] ) used isolated plosive-vowel syllables for which the preparatory visual movement is actually realized without any audio counterpart , hence producing a clear visual anticipation ( see Figure 5 ) . But the pervasive message linking visual lead and visual prediction within a predictive coding stance needs some refinement . Actually , as shown in the last part of this paper , audiovisual predictability does not require audiovisual asynchrony . The development of realistic computational proposals for assessing auditory and audiovisual prediction coding models in speech perception is a challenge for future work in cognitive neuroscience . For this perspective , precise knowledge of the natural statistics of audiovisual speech is a pre-requisite . A number of useful and important data and principles were provided in [21] , though the last of its four conclusions needed some refinement . The present paper hopefully contributed to enhance the available knowledge about the complexity of human speech . | Since a paper was published in this journal , an increasing number of neuroscience papers capitalize on the assumption that visual speech would be typically 150 ms ahead of auditory speech . It happens that the estimation of audiovisual asynchrony in the mentioned paper is valid only in very specific cases , for isolated consonant-vowel syllables or at the beginning of a speech utterance . But the view that vision leads audition is globally oversimplified and often wrong . It should be replaced by the acknowledgement that the temporal relationship between auditory and visual cues is complex , including a range of configurations more or less reflected by the temporal integration window from 30 to 50 ms auditory lead to 170 to 200 ms visual lead . This has important consequences for computational models of audiovisual speech processing in the human brain . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"linguistics",
"psychology",
"biology",
"and",
"life",
"sciences",
"social",
"sciences",
"speech"
] | 2014 | No, There Is No 150 ms Lead of Visual Speech on Auditory Speech, but a Range of Audiovisual Asynchronies Varying from Small Audio Lead to Large Audio Lag |
Katanin is an evolutionarily conserved microtubule-severing complex implicated in multiple aspects of microtubule dynamics . Katanin consists of a p60 severing enzyme and a p80 regulatory subunit . The p80 subunit is thought to regulate complex targeting and severing activity , but its precise role remains elusive . In lower-order species , the katanin complex has been shown to modulate mitotic and female meiotic spindle dynamics and flagella development . The in vivo function of katanin p80 in mammals is unknown . Here we show that katanin p80 is essential for male fertility . Specifically , through an analysis of a mouse loss-of-function allele ( the Taily line ) , we demonstrate that katanin p80 , most likely in association with p60 , has an essential role in male meiotic spindle assembly and dissolution and the removal of midbody microtubules and , thus , cytokinesis . Katanin p80 also controls the formation , function , and dissolution of a microtubule structure intimately involved in defining sperm head shaping and sperm tail formation , the manchette , and plays a role in the formation of axoneme microtubules . Perturbed katanin p80 function , as evidenced in the Taily mouse , results in male sterility characterized by decreased sperm production , sperm with abnormal head shape , and a virtual absence of progressive motility . Collectively these data demonstrate that katanin p80 serves an essential and evolutionarily conserved role in several aspects of male germ cell development .
The regulation of microtubule dynamics is an essential requirement for all cells and in many aspects of their daily function . The ability to precisely regulate microtubule number , the assembly of networks , and the rate of microtubule assembly and disassembly underpins cellular processes including division , differentiation and migration . Male gamete development in particular relies upon the co-ordinated development and rapid remodelling of complex microtubule structures , such as the mitotic ( spermatogonia ) and meiotic ( spermatocyte ) spindle; flagella formation needed for sperm motility; and the manchette , which determines sperm head shape and contributes to tail structure . Approximately one in 20 men of reproductive age is sub-fertile or sterile , of which 60% of cases are due to intrinsic defects in spermatogenesis . This heterogeneous disorder manifests clinically as diminished sperm number , or abnormal motility or morphology , or commonly combinations thereof , in the ejaculate [1] . All of these clinical presentations may be underpinned by defective microtubule dynamics . Microtubule severing is emerging as a key regulator of microtubule dynamics [2] , [3] , [4] , [5] , . The most well characterized microtubule severing enzyme is the katanin complex [8] , the severing function of which is carried out by an ATPase enzymatic subunit , named p60 , encoded by the Katna1 gene . Katanin p60 is a member of the AAA domain ( ATPases Associated with diverse cellular Activities ) protein family . Upon binding ATP , katanin p60 oligomerizes onto the tail of an individual tubulin subunit within a microtubule to form a 14–16 nm ring structure [9] . ATP hydrolysis confers a conformational change in the oligomer and ‘tugs’ upon the tail of the tubulin subunit . This leads to destabilization , and ultimately severing , of the polymer [2] . Other AAA microtubule-severing proteins include spastin and fidgetin [8] . Mutations in the gene encoding spastin cause progressive axon degeneration and underlie ∼40% of autosomal dominant cases of hereditary spastic paraplegia [10] and deletion of the fidgetin gene in mice results in a severe behavioural and developmental phenotype [11] , illustrating the importance of the family in neuronal development . The regulation and compartmentalization of microtubule severing is essential for normal cell function and survival . Katanin p60-mediated severing can be modulated by a p80 regulatory subunit , encoded by the Katnb1 gene [9] . The p80 subunit of katanin binds to p60 and targets it to the centrosome in transfected mammalian cell lines [9] , [12] , and generally enhances severing , but can also inhibit it depending on the cellular context [4] , [12] . In Tetrahymena thermophila [13] and Caenorhabditis elegans [14] , p80 null mutants phenocopy p60 null mutants . However the in vivo role of the p80 subunit in mammals remains enigmatic . Katanin was identified as a heterodimer of p60 and p80 subunits in sea urchins [8] , however the ratio of the two subunits shows developmental and regional variation in rat neurons [15] and in mouse testis ( the current study ) , suggesting that the expression of the p80 regulatory subunit may be one way in which p60-mediated severing is controlled in mammals [4] . The Katnb1 gene contains a C-terminal WD40 domain predicted to be involved in protein-protein interactions , and as such is a strong candidate for targeting p60-mediated severing to particular locations within a cell and for targeting p60-mediated severing to post-translationally modified or microtubule-associated protein ( MAP ) -associated tubulin polymers [16] , [17] . Katanin p80 also binds to the molecular motor protein dynein [18] and to dynein-regulating proteins [18] , [19] and could thus be involved in the transport of the katanin complex to specific sites . Katanin function is evolutionarily conserved , with p60 and p80 orthologues identified in species from all 5 kingdoms , including in C . elegans , Drosophila melanogaster , Arabidopsis thaliana , Chlamydomonas reinhardtii , mice and humans . Katanin localizes to mitotic spindle poles in mammalian cell lines , where it regulates spindle structure and chromosome movement [2] , [3] , [20] , [21] . Mutations in C . elegans orthologues of katanin p60 and p80 reveal roles for katanin in oocyte meiotic spindle assembly and chromosome movement [20] , [22] , and katanin regulates different microtubule populations , including kinetochore-associated bundles , to control oocyte meiotic spindle length in Xenopus [23] . Of note , katanin regulates mitotic chromosome movement in D . melanogaster by participating in the so-called “Pacman-mediated” shortening of spindle microtubule plus-ends which results in the poleward movement of the chromosomes [24] . In addition , mutations in katanin orthologues in two distantly related organisms Tetrahymena and Chlamydomonas result in the absence of the central axoneme microtubule pair and cilia/flagella defects , indicating a role in ciliogenesis [13] , [25] , [26] . To date there have been no in vivo models of katanin dysfunction in mammals . Here we show that a missense mutation in the highly conserved WD40 domain of the Katnb1 gene , encoding the p80 regulatory subunit of katanin , causes male sterility in mice characterized by oligoasthenoteratozoospermia . Katnb1 mutant mice , denoted as Taily , show frequent failure of meiotic spindle resolution , defective manchette function and abnormal axoneme development . Our findings highlight the critical role for katanin p80 in the regulation of microtubules dynamics in many aspects of male germ cell development . These data raise the possibility that defective katanin function may also contribute to human infertility , specifically defective sperm number , morphology and/or sperm motility .
Mouse lines carrying ENU-induced mutations causing male sterility were identified using breeding trials as described previously [27] . Lines with G3 male sterility at a frequency of one in four , but with normal mating behaviour , were chosen for further analysis . They included the ‘Taily’ line . The causal mutation was mapped using SNP-based methods and ultimately narrowed to a linkage interval on chromosome 8 ( between SNP markers rs3089148 and rs3710112 ) containing 74 genes . Candidate genes were chosen on the basis of testis expression and proposed function . The protein-coding region and intron-exon boundaries of 30 genes were sequenced . The causative mutation in Taily mice was identified as a recessive G to T substitution in exon 9 of the Katnb1 gene . No other mutations were found . Unaffected males possessed either homozygous wild type alleles ( Katnb1WT/WT ) or were heterozygous for the wild type and Taily allele ( Katnb1WT/Taily ) . Greater than 50 mice of each genotype were assessed and the genotype-phenotype correlation was absolute . The Taily mutation resulted in the conversion of a valine ( GTC ) to a phenylalanine ( TTC ) in the WD40 repeat region of the katanin p80 protein ( Figure 1 ) . The presence of an aliphatic amino acid ( e . g . V or I ) at amino 234 , relative to the mouse sequence , is absolutely conserved across all species , and is strongly suggestive of a functionally important role ( Figure 1B ) . Western blot analysis revealed that haploid germ cells from Katnb1Taily/Taily mice contained markedly less p80 protein than those from Katnb1WT/WT mice ( Figure 2 ) , demonstrating that the ‘Taily’ allele likely results in a loss-of-function . Katnb1Taily/Taily males showed no overt behavioural abnormalities , were morphologically identical to wild type littermates , were of normal weight ( Figure S1 ) , but were uniformly sterile when mated with wild type females ( n≥10 , Katnb1Taily/Taily males aged ≥8 weeks of age ) . Katnb1Taily/Taily females had apparently normal fertility . Testes from adult ( 8–12 weeks ) Katnb1Taily/Taily mice were 18 . 7% smaller than those from wild type littermates ( p<0 . 0001 , Figure 3A ) . Seminiferous tubules contained all germ cells types . Two major discordant features were apparent within the seminiferous epithelium from Katnb1Taily/Taily mice: 1 ) abnormally shaped spermatid heads ( Figure 3C and 3D ) and 2 ) abnormal meiotic cells at metaphase-anaphase ( Figure 3E and 3F ) . Stereological analysis revealed that the number of Sertoli cells , spermatogonia and spermatocytes per testis was not different between genotypes ( Figure 3B and Table S1 ) , indicating that the initiation of spermatogenesis and entry into meiosis were unaffected by the Taily mutation . The latter finding suggested that the function of the Sertoli cell blood-testis-barrier was normal , a proposition supported by the appearance of normal inter-Sertoli cell junctions by electron microscopy ( not shown ) . By contrast , testes from Katnb1Taily/Taily mice contained ∼30% fewer spermatids ( round and elongating ) compared to wild type ( Figure 3B ) . The reduction in spermatid populations was due to a decrease in the number of cells exiting meiosis , specifically during the final meiotic division in stage XII ( Figure 3B and Table S1 ) . TUNEL-labelling revealed that apoptotic cells were predominantly present in stage XII and stage I tubules which is when the final events of meiosis occur ( Figure 3G ) . Apoptotic spermatocytes in the process of meiotic division were observed ( Figure 3H ) . Collectively these results indicate that katanin p80 function is required for the final phases of male meiotic cell division . Stereology also showed that additional germ cells were not lost as they progressed through spermiogenesis ( Table S1 ) , however there was a 36 fold increase in the number of spermatozoa being phagocytosed by Sertoli cells ( Figure 3B ) in stages IX–XI tubules ( Figure 3I and 3J ) . These data indicate a failure in spermiation , the process by which sperm are released by the Sertoli cell at the end of their development , prior to their passage to the epididymis [28] . The data show that a significant proportion of spermatozoa failed to be released from the Sertoli cell and were instead phagocytosed , thus leading to a reduced number of sperm entering the epididymis . As a consequence , the epididymides from Katnb1Taily/Taily males contained a lower total number of sperm than would be anticipated from the testicular daily sperm output , i . e . 11% of wild type in the epididymis compared to 57% in the testis ( Figure 4A ) . Of the sperm found in the cauda epididymis of Taily mice , when compared to wildtype ( Figure 3K ) , all had abnormally shaped heads ( Figure 3L ) and displayed compromised total motility as assessed by computer assisted sperm analysis ( 80 . 3% in Katnb1WT/WT versus 37 . 2% in Katnb1Taily/Taily ) ( Figure 4B ) . Very few sperm were capable of forward ( progressive ) motility ( 52 . 4% in Katnb1WT/WT versus 11 . 1% in Katnb1Taily/Tail ) ( Figure 4B ) . Collectively these data indicate that katanin p80 has a role in germ cell exit from meiosis , in the establishment of structures or pathways within the sperm tail involved in motility and in the shaping of the sperm head . Katnb1Taily/Taily males were sterile as a consequence of decreased sperm production , abnormal sperm morphology and sperm being unable to ascend the female reproductive tract following mating . The analogous human phenotype is referred to as oligoasthenoteratozoospermia ( low sperm count , poor motility and abnormal shape ) . Both katanin p80 and p60 mRNAs were expressed in the testis during the post-natal establishment of the spermatogenic cycle ( Figure 5 ) . The katanin p60 microtubule severing enzyme was expressed at relatively similar levels in all ages examined , suggesting expression in multiple cell types ( Figure 5A ) . Katanin p80 regulatory subunit expression , however , peaked at day 30 , suggesting predominant expression in post-meiotic haploid spermatids ( Figure 5A ) . These data are consistent with previous microarray data ( germonline . org ) indicating that the katanin p60 catalytic subunit is expressed in Sertoli cells and germ cells to a similar degree , whereas the p80 regulatory subunit while detectable in Sertoli cells and spermatogonia , is more highly expressed ( 5-fold ) in spermatocytes and spermatids [29] . In accordance with the mRNA data , katanin p80 protein was most strongly localized within round through to elongating spermatids ( Figure 5B ) . Katanin p60 was also prominent in spermatids and was visible in Sertoli cells ( Figure 5B ) . In addition , the katanin p60 orthologues Katnal1 ( p60-like 1 ) and Katnal2 ( p60-like 2 ) were also expressed within the developing post-natal testis with a timing similar to that observed for katanin p80 ( Figure S2 ) . Absence of KATNAL1 immunolocalization in germ cells ( Smith et al , submitted for publication ) and immunolocalization of KATNAL2 predominantly to the sperm tail and cytoplasm , but not associated with the sperm head , ( Figure S2 ) , suggests that the meiotic and spermatid head-shape phenotypes reported herein were primarily mediated by katanin p80 regulation of the eponymous p60 subunit . The loss of germ cells during meiotic division prompted an investigation of the microtubule-based meiotic spindle in Katnb1Taily/Taily males . Compared to Katnb1WT/WT littermates , all Katnb1Taily/Taily meiotic spindles were abnormal ( Figure 6A and 6B ) . Specifically , metaphase spindles appeared to be more densely populated with microtubules , and projected from the poles at a wider angle than those from wild type animals ( Figure 6A and 6B and Videos S1 and S2 ) . Pole-to-pole measurements in metaphase spindles were longer in Katnb1Taily/Taily compared to Katnb1WT/WT ( 12 . 15±0 . 16 , n = 58 , versus 10 . 52±0 . 17 µm , n = 78 , mean ± SEM , p<0 . 0001 ) . Within metaphase and anaphase cells , p60 ( Figure 6C and Videos S1 and S2 ) and p80 ( Figure 6A and 6B ) proteins localized to microtubules of meiotic spindles . The Katnb1Taily/Taily mutation did not overtly alter this localization . Both p60 and p80 were observed along the microtubules of the spindle and at the microtubule-chromosome interface . The latter localization is consistent with katanin involvement in the poleward movement of chromosomes in Drosophila mitotic cells [24] . Specifically within Drosophila , katanin is believed to participate in the depolymerization of microtubule plus-ends in the midzone at anaphase during Drosophila mitosis , effectively “chewing away” the microtubule ends to facilitate spindle shortening via a process known as “Pacman” [24] . A role for katanin p80 in the Pacman-mediated poleward movement of chromosomes in mammalian meiotic anaphase is further supported by the appearance of multiple cells stalled in late anaphase in Katnb1Taily/Taily mice ( Figure 3F ) . Disordered meiosis is further evidenced by the frequent occurrence of binucleated haploid spermatids in Katnb1Taily/Taily testis sections ( Figure 6D ) . Binucleated spermatids were never observed in wild type animals . These data strongly suggest a role for katanin p80 in cytokinesis and midbody resolution . This hypothesis is supported by the localization of katanin p80 ( Figure 6E ) and p60 ( Figure S3 ) to the microtubules of the midbody in late telophase cells in both Katnb1WT/WT ( Figure 6E ) and Katnb1Taily/Taily germ cells ( Figures S3 and data not shown ) . In addition , telophase cells with prominent midbody microtubules were observed in testes from Katnb1Taily/Taily , but not in wild type animals ( Figure 6F ) . Taken together , the results suggest that katanin p80 , most likely in association with p60 , has a prominent role in midbody dissolution in male meiotic cells , and that this function is disrupted in Taily mice . Abnormal sperm head shape ( Figure 3D and 3L ) is frequently associated with defects in the function of the manchette [30] , [31] , [32] . The manchette is a transient microtubule structure assembled in elongating spermatids with proposed roles in both the sculpting of the sperm head and in the movement of proteins destined for the sperm tail , via a process referred to as intra-manchette transport ( IMT ) [33] . The manchette is comprised of large , parallel arrays of microtubule bundles that extend from beneath the acrosome/acroplaxome region of the spermatid head and project into the spermatid cytoplasmic lobe containing the growing sperm tail ( [34] and Figure S4 ) . The manchette is first seen at step 8 of spermiogenesis , when the round spermatid nucleus polarizes to one side of the cytoplasm , and the spermatid commences elongation . Nucleation of microtubules in the manchette is thought to occur on the perinuclear ring region of the spermatid head ( Figure S4 ) , and large parallel bundles are assembled as the spermatid nucleus starts to change shape in step 9 . In order to investigate the hypothesis that head abnormalities in sperm from Katnb1Taily/Taily mice were the consequence of abnormal manchette structure or function , testis sections were examined using electron microscopy . Manchettes in wild type elongating spermatids displayed the characteristic perinuclear ring and microtubule array structure ( Figure 3C ) . Those from Katnb1Taily/Taily elongating spermatids , however , displayed several abnormalities including constricted perinuclear rings , nuclear distortion and abnormally long microtubules extending into the distal cytoplasm ( Figure 3D ) . A stage-by-stage comparison of sections from wild type and Katnb1Taily/Taily males suggested defective manchette resolution in mutant animals . Although manchettes eventually resolved , the removal of manchettes in Katnb1Taily/Taily males was delayed . In wild type mice , manchettes normally reduce in size and then disappear in step 13 spermatids . In contrast , and when compared with wild type mice ( Figure 7A ) , in step 13 spermatids from Katnb1Taily/Taily mice , abnormally long manchette microtubules extended into the cytoplasm and were associated with tubulin-labelled ‘clouds’ ( Figure 7B ) . The timing and location of the ‘clouds’ is suggestive of abnormal microtubule disassembly . The above defects were confirmed and more dynamically visualized in elongating spermatids isolated from wild type and Katnb1Taily/Taily males labelled with α-tubulin and TOPRO to visualize microtubules and the nucleus , respectively ( Figure 8 and Videos S3 and S4 ) . An analysis of progressively more mature elongating spermatids revealed that while manchettes in Katnb1Taily/Taily mice appeared to form at the correct time and initially began to move distally as spermiogenesis proceeded , movement stalled at approximately step 10 ( Figure 8 ) . By contrast , the progressive constriction of the peri-nuclear ring that normally occurs as the manchette moves over the caudal half of the spermatid , continued to occur ( Figure 8 ) . This resulted in a bulbous nuclear shape forward of the stalled peri-nuclear ring , and an abnormally elongated nucleus distally , resulting in the unusual ‘knob-like’ head structure also visible at an electron microscopic level ( Figure 3D ) . As observed by electron microscopy ( Figure 3D ) and in testis sections ( Figure 7B ) abnormal elongated manchette microtubules were easily observed in isolated late step spermatids ( Figure 8 ) . Consistent with the defects seen in Katnb1Taily/Taily mice , both the katanin p80 regulatory subunit and the p60 catalytic subunit localized to microtubules of the manchette ( Figure 7C and 7D ) . Focal labelling of katanin subunits was observed along the manchette microtubules and particularly at the microtubule ends projecting into the cytoplasm ( Figure 7C and 7D ) . This localization is consistent with a role for katanin-mediated severing in regulating manchette length . Localization was not obviously affected in the manchette of Katnb1Taily/Taily mice ( not shown ) . Both subunits also localized to the acrosome/acroplaxome region , in a manner reminiscent of proteins that undergo trafficking in an acrosome/acroplaxome-manchette-flagella pathway [35] . Collectively these results reveal that katanin p80 , most likely in association with p60 , has an essential role in both the formation of the manchette and in the dynamics of its movement and resolution . Perturbed katanin p80 function results in the failure of manchette migration , abnormally long and mal-orientated manchette microtubules and abnormal dissolution . Such abnormalities are entirely consistent with the observed defects in sperm head shape ( teratozoospermia ) in sperm from Katnb1Taily/Taily animals . The manchette also plays a critical role in the development of sperm flagella via a process known as intra-manchette transport , or IMT [33] . IMT is thought to be involved in sperm tail development in a manner analogous to intra-flagella transport in somatic cilia and in flagella in organisms including Chlamydomas and Trypanosomatids [36] . Defects in manchettes and sperm motility in Katnb1Taily/Taily mice , and a previously demonstrated role for katanin p80-mediated ( PF15p ) severing in the formation of axoneme central microtubules in Chlamydomonas [26] , prompted us to investigate flagella/tail structure in Katnb1WT/WT and Katnb1Taily/Taily sperm . When compared to controls ( Figure 7E ) , electron microscopic analysis revealed a variety of axoneme defects , including missing central microtubules and hemi-axonemes ( Figure 7F ) . Of note , the majority of Katnb1Taily/Taily sperm also contained flimsy or missing outer dense fibers ( Figure 7F ) consistent with previously proposed roles for the manchette in the transport of proteins into the developing sperm tail and formation of accessory tail structures [33] . The outer dense fibers are rod-like structures running parallel to , and connected to , the microtubules of the axoneme that are believed to protect sperm against shearing forces and to provide directionality to tail bending ( reviewed in [37] ) . Collectively , these defects demonstrate an essential role for katanin p80 in the development of the sperm flagellum , including the axoneme and the formation of the accessory structures .
Analysis of male mice with a mutation in the regulatory subunit of the katanin microtubule severing enzyme complex revealed several roles for katanin p80 in mammalian microtubule dynamics . These studies reveal an essential requirement for katanin p80 in male fertility , and in multiple aspects of mammalian male gamete development , including in meiotic spindle dynamics , cytokinesis , flagella development and sperm head shaping . Together with in vitro data and evidence of disordered microtubule structure and function in lower order species , this novel mouse mutation reveals that katanin has roles in controlling microtubule dynamics . While severing can result in microtubule destruction , it is also important for the creation of new microtubules , via the severing of longer stable microtubules into shorter segments that can then be used as “seeds” for further microtubule polymerization [4] , [7] . For example , katanin severs newly created microtubules at the neuronal centrosome , facilitating their transport to other sites , such as within developing axons [38] . Katanin can sever microtubule lattice defects in a quality control mechanism [39] , and in neurons can create branched microtubule networks [4] , [40] . Recent data also revealed that katanin can depolymerize microtubules at their plus-ends [5] , [16] , [39] and finally , the p60-p80 katanin complex has a severing-independent , microtubule cross-linking function at C . elegans oocyte meiotic spindle poles [23] . The p80 protein contains a WD40 domain that likely mediates protein-protein interactions [9] , [12] . The C-terminal region interacts with the p60 enzyme [12] and contains binding sites for the molecular motor protein dynein . The C-terminal region also binds to the dynein-associated proteins LIS1 and NDE1 in neurons [18] . While p80 is thought to modulate p60 targeting and activity [3] , [9] , [12] , the precise in vivo roles of p80 are not well understood . The Katnb1Taily/Taily mutation in the WD40 domain results in decreased p80 protein within germ cells and defects in microtubule-based processes . Based on the position of the mutation , and on the fact that less p80 protein is produced in mutant mice , we predict that the mutation influences the ability of p60 to sever , as well as the targeting of this severing activity to specific sites within the cell . This mouse model recapitulates many of the proposed functions of katanin observed in lower order species . It is of note , that the manchette defects observed in Taily mice phenocopy many of the defects observed in a Lis1 null mice [31] . This observation and previous studies showing the localization of LIS1 and NDE1 in the manchette [41] , suggests that similar to the proposed role for these proteins in neurons , the katanin complex co-operates with LIS1 and NDE1 during sperm head shaping . This interaction will be the subject of future investigations . The data demonstrate a role for katanin p80 in mammalian male meiotic cell division . Null mutations in the C . elegans p80 ortholog mei2 are associated with meiotic defects in oocytes , including an inability to assemble a meiotic spindle [14] . Katanin function in male germ cells has not previously been studied to the best of our knowledge . Our observations on male metaphase spindles in Katnb1Taily/Taily mice are consistent with the longer metaphase meiotic spindles produced in C . elegans oocytes with a partial loss-of-function mei2 mutation [20] and with the recent demonstration of a conserved role for katanin in controlling the length of meiotic metaphase spindles in Xenopus oocytes [23] . Within C . elegans , the p80 protein targets the p60 severing enzyme to the spindle poles in meiotic oocytes [20] , [42] . We did not observe either p60 or p80 at spindle poles in male meiotic germ cells . All meiotic spindles were , however , abnormal , indicating a role for katanin in the assembly of spindles , as supported by various studies [14] , [20] , [21] , [23] . In meiotic male germ cells , the most obvious localization of katanin subunits was to the microtubule ends near the chromosomes in metaphase-anaphase cells , suggesting a role for katanin in the shortening of microtubule plus ends during anaphase . The “Pacman-mediated” shortening of microtubule plus-ends within the spindle midzone is important for the poleward movement of chromosomes in mitotic anaphase [24] , but has not been studied in meiotic cells . Our observation of cells apparently stalled in anaphase , together with the finding that 30% of cells die during the later phases of meiosis , supports the hypothesis that disturbed p80 function causes defects in the poleward movement of chromosomes during anaphase . Such defects result in disturbed spindle resolution and , often , cell death . Finally , the appearance of binucleated spermatids and the localization of katanin subunits to the midbody in meiotic cells in male mice supports the hypothesis that p80 , and potentially the katanin complex , has a conserved role in modulating microtubule dynamics at the midbody during meiotic cytokinesis . In support , katanin p80 dysfunction or mislocalization is associated with defective mitotic cytokinesis in Trypanosomes [25] and in sarcoma cells in vitro [43] . Katanin p80 is essential for sperm head shaping via the regulation of the manchette , which is in itself a complex microtubule network . An analysis of manchette position during spermiogenesis indicated that manchette movement is defective in Katnb1Taily/Taily mice , suggesting katanin is involved in the organization and remodelling of this microtubule network as it moves over the nucleus . The localization of p60 and p80 within the manchette , together with the Taily phenotype is consistent with a role for katanin action at multiple sites . These include 1 ) the severing of microtubules at the perinuclear ring , thereby facilitating the release of microtubules from the nucleating center , and the production of the microtubule lattice , as has been proposed in other systems [4]; 2 ) the severing of microtubules near the nucleus to permit movement of the manchette perinuclear ring as it shapes the nucleus; 3 ) within the microtubule lattice to facilitate remodelling of this complex structure; and 4 ) the severing of microtubules at the caudal end of the manchette to control manchette length and dissolution . Katanin activity also regulates the dynamics of large microtubule-based array structures in neurons [18] , [38] , [40] . Taken together , the data support the hypothesis that p80 , and katanin function , is important for the movement and remodelling of large microtubule arrays in mammalian cells . This role is essential for the normal development and shaping of sperm , which in turn is critical for normal sperm function and male fertility . Finally , we demonstrate for the first time that katanin p80 is required for mammalian sperm flagella development and subsequent motility . A conserved role for katanin in the assembly and disassembly of cilia and flagella has been revealed in two distantly related lower order species Chlamydomonas [26] and Tetrahymena [13] ( reviewed in [6] ) . Katanin activity controls flagellum length in Trypanosomatids [25] and severs axonemal microtubules during the deflagellation process in Chlamyodmonas [44] , [45] . In Chlamydomonas , katanin p80 is also specifically required for the assembly of the central microtubule doublet of the flagellum axoneme [26] . Defects in axonemal structures ( including missing central microtubule doublets ) and outer-dense fibers in Katnb1Taily/Taily mice suggest that katanin p80 regulates sperm motility by acting at multiple sites in sperm development . Specifically , p80 function is required in the regulation of axonemal assembly and in the delivery of proteins to the developing flagellum via IMT . Given the high level of expression of katanin p80 in other tissues , and the proven role for katanin in C . elegans oogenesis , it is surprising that other overt phenotypes were not noted in mutant mice . We hypothesize that other phenotypes will be revealed when mice are exposed to environmental insults . Studies into the role of katanin in oocyte function are ongoing . In conclusion , the p80 subunit of the katanin microtubule severing enzyme complex is required for male fertility in mice . This is the first in vivo mammalian model of katanin p80 dysfunction , and it presents with a phenotype reminiscent of a commonly observed clinical phenotype of male infertility characterized by low sperm counts , poor motility and abnormal sperm morphology ( referred to as oligoasthenoteratospermia or OAT ) . We conclude that p80 katanin is required for male meiotic spindle development and dynamics , and for the shaping of the sperm head via the regulation of manchette development and movement . Katanin p80 also participates in meiotic cytokinesis , likely via the regulation of the microtubules within the midbody , and controls the development and function of sperm flagella .
All animal experimentation was approved by the Australian National University and Monash University Animal Experimentation Ethics Committees and performed in accordance with Australian NHMRC Guidelines on Ethics in Animal Experimentation . Point mutant mice were generated as described previously on a C57BL/6 background and outbred to CBA [27] . Mouse lines containing sterility causing mutations were identified by breeding trials wherein eight G3 brother-sister pairs per line were co-housed and the presence of pups monitored . Lines where male sterility was observed in a ratio of approximately one in four in the G3 generation with apparently normal mating behaviour were selected for further analysis . Affymetrix 5K mouse SNP Chip arrays were used to map the sterility causing mutation . Genomic DNA from five affected males was hybridized onto the array at the Australian Genome Research Facility and compared to wild type C57BL/6 and CBA sequences . The linkage interval was subsequently narrowed using additional mice and SNPs ( www . well . ox . ac . uk/mouse/INBREDS/ ) using the Amplifluor SNP Genotyping System ( Chemicon ) . Plates were read in a BMG Fluostar optima fluorescent microplate reader . Following the identification of the causal mutation , mice were specifically genotyped using the Amplifluor SNPs HT genotyping system using a wild type-specific antisense primer 5′-GAAGGTCGGAGTCAACGGATTAAGAGCACCCGTACCTGAC-3′ , a mutant allele antisense primer 5′-GAAGGTGACCAAGTTCATGCTGAAGAGCACCCGTACCTGAA-3′ , a sense primer , 5′-GGTGGTGAGCTGCATTGAA-3′ and Platinum Taq DNA Polymerase ( Invitrogen ) . Conditions for amplification were as follows: 4 minute denaturation at 95°C , 35 cycles of denaturation at 95°C for 10 seconds , annealing at 60°C for 20 seconds and elongation at 72°C for 40 seconds , followed by a final 3 minute elongation at 72°C . Following the reaction , plates were read in a BMG Fluostar optima fluorescent plate reader . Infertility in the Taily mouse line was classified using the scheme outlined in Borg et al [46] . Daily sperm output and total epididymal sperm content were determined as described previously [47] . Sperm motility was assessed using computer assisted sperm analysis [48] and ultra-structure using electron microscopy [49] . Cauda epididymal sperm morphology was assessed following staining with hematoxylin . Cells undergoing apoptosis were visualized using the Apoptag kit ( Millipore ) as recommended by the manufacturer . The number of germ cells per Sertoli cell were enumerated in 25 µm thick , periodic acid Schiffs ( PAS ) stained methacrylate sections using the optical disector as previously described [50] . Retained elongated spermatids were counted in stage XI-XI [50] and expressed as fold wild type . RNA was extracted from testes at defined periods throughout post-natal development using TRIzol regents ( Life Technologies ) , treated with DNase I ( Ambion ) and cDNA sythnesized using oligo-dT primers and SuperScript III reverse transcriptase ( Life Technologies ) . The relative expression of Katnb1 , Katna1 , Katnal1 and Katnal2 were defined using quantitative PCR using TaqMan assays ( Applied Biosystems ) Mm01244795_m1 , Mm00496172_m1 , Mm00463780_m1 and Mm00510701_m1 respectively . Expression of these was normalized against peptidylprolyl isomerase A ( Mm002342429_g1 ) . Germ cell sub-populations were purified using the Staput method as previously described [51] . Single cell suspensions were loaded onto a 2–4% continuous BSA gradient and elongated spermatids and round spermatids collected after a 3 hour and 3 . 5 hour sedimentation period , respectively . For immunofluorescent staining , gradient fractions were pelleted and resuspended in 4% paraformaldehyde ( PFA ) fixative for 2 hours on ice . Cells were then washed with PBS and spread onto slides . Protein was extracted from round spermatid fractions ( >90% purity ) using 20 µL M-PER buffer ( Thermo Scientific ) . 10 µg of protein was separated on a 10% SDS-PAGE gel and probed for rabbit katanin p80 ( HPA041165 , which recognizes a C-terminal region of p80 , Sigma Aldrich ) and actin ( A2066 , Sigma Aldrich ) . Bound antibody was detected using a goat anti-rabbit IgG HRP ( P0488 , Dako ) secondary antibody and an enhanced chemiluminescence ( ECL Plus ) detection kit ( Amersham Biosciences ) . Katanin subunits and α-tubulin were localized in testis sections as described [52] . Primary antibodies included: anti-α-tubulin ( T5168 , Sigma , diluted 1 in 5000 ) , anti-katanin p60-like 2 ( p60AL2 , #sc-84855 , Santa Cruz , diluted 1 in 100 ) , anti-katanin p80 ( diluted 1 in 200 ) [9] and anti-katanin p60 ( diluted 1 in 200 ) [53] . Both p60 and p80 antibodies were affinity-purified from rabbits immunised against full length recombinant human proteins . These antibodies have been validated extensively and have been shown to recognize a single polypeptide in HeLa cells and a range of human tissues [9] , [53] . Given the sequence homology between the p60 and p60L1 subunits , and between p80 and the uncharacterized c15orf29 subunit , there remains the possibility of partial cross-reactivity . Secondary antibodies included: Alexa Fluor 555 donkey anti-rabbit IgG ( A-31572 ) and Alexa Fluor 488 donkey anti-mouse IgG ( A-21202 ) ( diluted 1 in 500 ) . DNA was labelled using DAPI ( Invitrogen ) . To define the localization of proteins within isolated elongating spermatids , cells were permeabilized in 0 . 2% Triton X-100 diluted in 10% normal horse serum ( NHS ) in PBS for one hour at room temperature . Non-specific labelling was minimized by blocking in 10% NHS in PBS for 30 minutes . Primary antibodies were diluted in 10% NHS in PBS and incubated overnight at 4°C . Secondary antibodies were diluted 1 in 200 and incubated at room temperature for 2 hours . DNA was labelled using TOPRO3 ( Invitrogen , 1 in 200 ) or DAPI . Images were taken with an SP5 5-channel ( Leica Microsystems ) confocal microscope in the Monash University Microimaging facility . Metaphase spindle lengths were measured on α-tubulin and DAPI-stained sections from Katnb1WT/WT and Katnb1Taily/Taily mice using LAS AF ( Leica Application Suite Advanced Fluorescence ) software . Z-stacks of spindles and manchettes were collected at 0 . 5 µm intervals . Images were assembled using Adobe Photoshop . Test and subject images were adjusted uniformly across the image and between groups . Differences between Katnb1WT/WT and Katnb1Taily/Taily mice were determined using unpaired t tests in GraphPad Prism 5 . 0 . | Microtubules are critical components of cells , acting as a “scaffold” for the movement of organelles and proteins within the cytoplasm . The control of microtubule length , number , and movement is essential for many cellular processes , including division , architecture , and migration . We have defined the role of the microtubule severing protein katanin p80 in male germ cell development . Male mice carrying a point mutation in the p80 gene are sterile as a consequence of low numbers of sperm , abnormal sperm morphology , and poor motility ( ability to “swim” ) . We show that this mutation is associated with defects in microtubule structures involved in the division of immature sperm cells , in structures that shape the sperm head , and in the sperm tail , which is essential for sperm movement in the female reproductive tract . This study is the first to show that katanin p80 , via its effects on microtubule dynamics within the testis , is required for male fertility . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"urology",
"developmental",
"biology",
"model",
"organisms",
"biology"
] | 2012 | An Essential Role for Katanin p80 and Microtubule Severing in Male Gamete Production |
Erythropoiesis is one of the best understood examples of cellular differentiation . Morphologically , erythroid differentiation proceeds in a nearly identical fashion between humans and mice , but recent evidence has shown that networks of gene expression governing this process are divergent between species . We undertook a systematic comparative analysis of six histone modifications and four transcriptional master regulators in primary proerythroblasts and erythroid cell lines to better understand the underlying basis of these transcriptional differences . Our analyses suggest that while chromatin structure across orthologous promoters is strongly conserved , subtle differences are associated with transcriptional divergence between species . Many transcription factor ( TF ) occupancy sites were poorly conserved across species ( ∼25% for GATA1 , TAL1 , and NFE2 ) but were more conserved between proerythroblasts and cell lines derived from the same species . We found that certain cis-regulatory modules co-occupied by GATA1 , TAL1 , and KLF1 are under strict evolutionary constraint and localize to genes necessary for erythroid cell identity . More generally , we show that conserved TF occupancy sites are indicative of active regulatory regions and strong gene expression that is sustained during maturation . Our results suggest that evolutionary turnover of TF binding sites associates with changes in the underlying chromatin structure , driving transcriptional divergence . We provide examples of how this framework can be applied to understand epigenomic variation in specific regulatory regions , such as the β-globin gene locus . Our findings have important implications for understanding epigenomic changes that mediate variation in cellular differentiation across species , while also providing a valuable resource for studies of hematopoiesis .
Red blood cell ( RBC ) production ( erythropoiesis ) is one of the best understood examples of lineage commitment and cellular differentiation [1]–[4] . This process begins as multipotent hematopoietic stem cells ( HSCs ) differentiate into lineage committed erythroid progenitors , losing multipotency in intermediate progenitor cell populations . Early erythroid progenitors then differentiate into morphologically distinct early erythroid precursors , termed proerythroblasts ( ProEs ) . The ProEs subsequently undergo terminal erythroid differentiation into mature RBCs that enucleate , contain a significant concentration of hemoglobin , and have highly elastic cytoskeletons [3] . This differentiation process is governed by a number of transcription factors ( TFs ) that dynamically coordinate a complex transcriptional gene regulatory network ( GRN ) . Importantly , much of our knowledge of this GRN has been derived from mouse models of erythropoiesis [1]–[3] . Extrapolation from mouse models of terminal erythroid differentiation to humans has historically been straightforward , grounded in the nearly identical morphology of mature RBCs and their precursors between species [4]–[6] . While there are many well-known examples of species-specific differences in erythroid GRNs , such as developmental variation of β-like globin gene expression , the divergent role of BCL11A during developmental hemoglobin switching , and differences in cis-regulatory modules ( CRMs ) regulating GATA1 transcription [7]–[9] , a marked global divergence in the expression profiles of the erythroid lineage was only recently described by systematic comparative analyses of human and murine erythroid transcriptomes [10] , [11] . Indeed , these recent studies independently identified a large global divergence in temporal patterns of gene expression between human and mouse at critical , canonical stages of terminal erythroid differentiation [10] , [11] . While many erythroid specific pathways and genes were generally conserved , such as the heme biosynthetic pathway , cytoskeletal proteins , and master TFs of erythropoiesis ( e . g . GATA1 , NFE2 , and KLF1 ) , significant differences in the timing and expression levels of certain constituent genes were observed ( e . g . TAL1 ) [10] . In some pathways , such as the mitogen-associated protein kinase ( MAPK ) pathway , gene profiles were markedly divergent between species during differentiation [11] . These differences have many important implications for integrating the extensive information on erythropoiesis gained from mouse models to better understand human erythropoiesis and how this process goes awry in human disease . For example , congenital dyserythropoietic anemia type II ( CDA II ) is caused by recessive mutations in SEC23B , but the phenotype could not be recapitulated in mouse models [12]–[14] . The expression of SEC23A , a SEC23B paralog , varied between mice and humans , suggesting a reason for these divergent phenotypes . Moreover , these expression differences were accompanied by variation in TF occupancy proximal to SEC23A in erythroid cell lines suggesting that species-specific differences in transcription may be due to evolutionary divergence in TF occupancy and the epigenome [10] . However , the conservation or divergence of chromatin structure and TF occupancy between human and murine erythropoiesis has only been characterized in a few specific regions , and , to the best of our knowledge , we are not aware of any studies that measure the extent to which there is divergence or conservation across the genome [7] , [12] . We have therefore undertaken a comparative epigenomic study to systematically analyze the global conservation of histone modifications and master transcriptional regulators necessary for erythroid differentiation . We map these epigenomic marks in both human and murine primary ProEs as well as in the model erythroid cell lines of human and mouse , K562 and G1E/G1E-ER ( herein referred to as G1E ) , respectively . We compare these marks in the context of orthologous genes as well as across conserved regions of both genomes . Finally , we integrate high-quality stage-matched gene expression profiling ( RNA-seq ) of each cell type to investigate functional intra- and inter-species differences across the epigenome . Our results suggest that chromatin structure and function is generally well conserved both between species and in erythroid cell models , although certain modifications are under greater constraint than others . In contrast , only ∼25% of the occupancy sites of most TFs are conserved between species , whereas we observed a 2-fold increase in conservation rates for erythroid cell models , validating K562 and G1E cell lines as species-specific model systems for studying such TFs . Nevertheless , we find that CRMs co-occupied by KLF1 , GATA1 , and TAL1 are significantly more conserved than any lower order combination of these factors and are strictly localized near highly-expressed genes that play a key role in defining erythroid cell state , suggesting that these regions are under strong evolutionary constraint to regulate common features of mammalian erythropoiesis . Moreover , although we show that chromatin structure is largely conserved between similar developmental cell-types across species , subtle changes in chromatin structure are associated with transcriptional divergence . Based on multiple lines of evidence , we suggest that evolutionary changes in transcription are partially driven by large-scale loss or gain of master TF occupancy that associate with changes to the underlying chromatin structure . In addition , these results provide a resource that can aid in translating findings from mouse erythropoiesis to the analogous process in humans .
For each species , we compiled chromatin immunoprecipitation high-throughput sequencing ( ChIP-seq ) data sets of histone modifications ( H3K4me1 , H3K4me2 , H3K4me3 , H3K9ac , H3K27me3 , H3K36me3 ) and master TFs of erythropoiesis ( GATA1 , TAL1 , KLF1 , NFE2 ) at the ProE stage of erythroid differentiation ( S1 and S2 Table ) [15]–[25] . The vast majority of ChIP-seq data was available at the ProE stage , and this is known to be an important time point where a variety of epigenetic changes occur to mediate alterations in the transcriptional landscape [19] , [20] , [26] , [27] . Additionally , we compiled and analyzed ChIP-seq data from erythroid cell lines , K562 ( human leukemia cell line ) and G1E/G1E-ER ( mouse erythroid cell lines that are derived from Gata1-null erythroid cells containing an estrogen-inducible Gata1 transgene; herein G1E ) . We initially leveraged the compiled data to investigate local chromatin structure and TF occupancy across 15 , 506 orthologous gene bodies with a one-to-one mapping because local chromatin structure is largely indicative of transcription status and interspecies TF occupancy differences [28] , [29] . Overall , our observations are concordant with prior data suggesting that the functions of histone modifications , indicated by similar histone intensity profiles and the percent of genes present near each , are well conserved between humans and mice ( Fig . 1 left panel , S1 & S2 Fig . ) [30] . For example , the signal intensity of H3K4me3 , generally regarded as a mark of transcriptional activation , was present in ∼50% of genes in both species and its intensity peaked at the transcription start site TSS ( Fig . 1A ) , while the pattern of H3K27me3 , a mark of transcriptional repression , was conserved overall but was present in a lower number of genes ( ∼20% ) ( Fig . 1C , S2 Fig . ) . When we compared TF ( GATA1 , KLF1 , TAL1 , NFE2 ) occupancy profiles across gene-bodies identical to the above analysis of histone modifications , we discovered that for each TF , normalized occupancy intensities varied significantly more between species than histone modifications ( S3 Fig . ) . One hypothesis for this observation is that certain TFs such as TAL1 are not as abundant or active in mouse versus human ProEs , although our expression data suggests that TAL1 is highly abundant at this stage in both species [11] . More likely hypotheses are that technical differences in ChIP protocol between labs explain most of the observed difference or that the differences are truly biological . A thorough analysis supporting these alternative hypotheses is detailed in the materials and methods . To quantify the potential divergence in chromatin structure and TF occupancy , we compared relative histone modification intensity across the proximal promoter regions of a smaller set of 6596 orthologous genes with canonical transcripts in both species . We included erythroid cell lines in this analysis and used human ProEs as the primary cell type against which all others were compared to assess inter- and intra-species conservation of promoter epigenetic structure . Generally , histone promoter modifications were highly conserved between the two species and replicate experiments were highly correlated ( middle and right columns of Fig . 1 , S3 Table ) . H3K4me3 , H3K4me2 , and H3K9ac were highly conserved between all intra- and inter-species cell types , although modifications in K562 were most correlated with those in human ProEs ( Fig . 1A , B , D ) . Interestingly , H3K27me3 was more conserved in mouse ProEs than K562 cells ( Fig . 1C ) . The observed divergence of H3K27me3 in the leukemic K562 cell line is consistent with the fact that H3K27me3 modifications are frequently dysregulated during oncogenesis [31] . H3K36me3 and H3K4me1 were moderately conserved and more strongly correlated between cell types than between species , although these marks show the weakest enrichment at the TSS ( S1 Fig . and S3 Table ) . When we compared TF occupancy intensity , GATA1 and TAL1 intensity in human ProEs was moderately correlated with that of K562 cells , but not with mouse ProEs or G1E cells ( S4 Fig . ) . In contrast , KLF1 and NFE2 occupancy was weakly to moderately correlated across all cell-types ( S4 Fig . ) . Importantly , K562 cells proved a significantly better model of promoter TF activity in comparison with primary human erythroid promoters than mouse ProEs . Comparing erythroid cell lines directly , the two classes of cell types showed the weakest correlation , consistent with their respective derivation from primary cells ( S3 Table ) . G1E cells showed similar correlations to mouse ProEs and were moderately to strongly correlated with mouse ProEs across all modifications ( S3 Table ) . Interestingly , these results suggest that active promoters , marked by H3K4me3 , H3K4me2 , and H3K9ac , are under strict evolutionary constraint and that conservation of these histone modifications is necessary for transcription that defines cell state across species . Overall , these data add to the increasing evidence that inter-species epigenetic differences are larger than intra-species differences – at least for cells that take on a similar global cellular state [29] . Promoters are only one piece of the total regulatory landscape , so we extended our analysis and performed a global cross-species comparison of chromatin structure and master TF occupancy to better understand patterns in epigenomic evolution . We investigated conservation of global occupancy patterns for all four master regulators of erythropoiesis . Briefly , we derived robust TF occupancy peaks and lifted narrow summits from the mouse genome to the human genome to assess conservation . We note that in this section , when we discuss conservation , we are primarily referring to “conservation of TF occupancy sites” between species or cell types . During the 75 million years of evolution separating the two species , ∼75% of master regulator ( GATA1 , TAL1 , and NFE2 ) occupancy sites were lost between humans and mice ( Fig . 2A , “mapped” ) . In stark contrast , greater than 60% of KLF1 occupancy peaks were conserved between species . Interestingly , we observed that in ∼25% of lost TF peaks , new human-specific occupancy sites were created for each TF in nearby regions ( +/- 5 kbs , henceforth known as “compensatory” occupancy sites ) , a phenomenon that has been described between human and mouse hepatocytes , adipocytes , and closely related Drosophila species ( Fig . 2A , “mapped +/- 5 kb” ) [32]–[34] . Nevertheless , although we observed that large numbers of TF occupancy sites were lost between species , each master regulator is far more conserved than expected by chance ( p<10−5 for each , permutation test ) and canonical TF binding motifs were nearly identical for each TF across species ( Fig . 2D ) . These findings suggest that the exact genomic location of each TF occupancy site may not be as functionally important as its presence in a broader genomic region and highlight the idea that some presumed cis-regulatory modules ( CRMs ) may have at most small functional effects . We also considered the differences in peaks called between species by mapping human peaks to mouse peaks ( Fig . 2B ) . We observed a similar ranking of TF conservation , although the percentages were overall much lower , reflective of the greater number of occupancy sites called in humans . These percentages represent a lower bound on the true percentage of conserved peaks , while those shown in Fig . 2B are a better estimate of the true conservation of TF occupancy rate . As a sensitivity analysis , we investigated conservation of only the strongest 25% of TF occupancy peaks , providing an upper bound on the conservation estimate for each TF ( S5 Fig . ) . In juxtaposition to mouse sites , TF occupancy sites in K562 cells and human ProEs were highly concordant: ∼50% of occupancy sites were identical , and only a small percentage of compensatory peaks were observed ( Fig . 2C ) . Importantly , the upper bound of the conservation estimates for human and mouse ProEs are still below the standard estimates for human ProEs and K562 cells . Overall , these data suggest that while select master regulators , such as KLF1 , are under strong constraint , most master regulators , including GATA1 , TAL1 , and NFE2 , are under weak to moderate constraint . Second , these data suggest that although TF occupancy sites are often lost during evolution , functional effects from these losses are partially buffered by the emergence of compensatory occupancy sites , a possibility that we validate in subsequent analyses . Our findings , both globally across the genome and in promoter regions , support the idea that intra-species TF occupancy is more conserved than inter-species TF occupancy . In contrast to GATA1 , TAL1 , and NFE2 binding motifs , KLF1 motifs ( SP1 or CACC ) were not centrally enriched around the summit of KLF1 occupancy sites ( although enrichment for the canonical motifs were observed across the entire peak ) . A thorough investigation of enriched motifs in KLF1 occupancy sites revealed that GATA1 and GATA1/TAL1 motifs were proximally , but not centrally , enriched in KLF1 peaks in human ( Fig . 2D , Fig . 3A–B , S1–S8 Dataset ) . Furthermore , KLF1 motifs were recovered in both GATA1 and TAL1 peaks for both human and mouse , suggesting that regions co-occupied by KLF1 , GATA1 , and TAL1 are true CRMs under stricter evolutionary constraint than regions occupied by each factor alone ( Fig . 3A ) . To address this hypothesis , we mapped combinatorial occupancy regions of these three factors from the mouse to the human genome ( S6A Fig . ) . We discovered that when one or more TF overlapped , the region was more likely to be conserved ( p<10−5 for each , Fig . 3D ) . Confirming our hypothesis , ∼35% of regions co-occupied by GATA1 , KLF1 , and TAL1 in mice were also co-occupied by all three factors in humans , a result far more likely than by chance and a higher rate of conservation than any other grouping ( p<10−5 , Fig . 3C–D ) . Our observation that certain co-occupied TFs are more conserved than individual TFs is consistent with similar findings across closely related mammalian species [35] . This result suggests that CRMs co-occupied by all three master regulators are important for the regulation of highly conserved processes during erythropoiesis . Confirming this , we found that the majority of these regions localize to and may act as enhancer elements for a number of genes important for erythropoiesis including: β-globin , heme biosynthetic enzymes , red cell membrane and surface proteins , and master regulators of erythropoiesis ( Fig . 3E–F ) . Additionally , the importance to erythropoiesis of many of the genes proximal to these constrained enhancers is unknown , providing a short list of potential new regulators of erythropoiesis under strict evolutionary constraint ( S4 Table ) . Overall , these observations validate the full extent to which KLF1 , in conjunction with GATA1 and TAL1 , regulates many facets of mouse and human erythropoiesis [18] , [36] , [37] . Considering the divergence in TF occupancy sites between species , we investigated the extent to which underlying chromatin structure was associated with the observed loss or gain of different master TF occupancy sites . We undertook a comprehensive approach to annotate all regions of the genome by utilizing a hidden Markov model ( HMM ) to infer 15 biologically meaningful chromatin “states” in ProEs and for K562 cells , each comprised of multiple different histone modifications with varying “strengths” ( i . e . frequencies ) for every 200 bp region across both genomes ( Fig . 4A , S7 Fig . , S8 Fig . , S9 Fig . , S5 Table , see materials and methods for details ) [38] . To facilitate comparisons across species , master regulator occupancy sites were grouped according to conservation . “Conserved” occupancy sites were defined as occupancy sites present in both mouse and human orthologous genomic regions , “lost” or “mouse-specific” were present in mouse but not in human , “compensatory” were gained in human proximal to a lost occupancy site , “gained” or “human-specific” were present in human but not in mouse , and “strongly gained” sites are the top 10% of human-specific occupancy sites . We observed that conserved occupancy sites were most significantly enriched for active chromatin states that include strong enhancers and promoters ( state 5 , 6 , 8 , 10 , 11 , 12 ) ( Fig . 4B ) . These states were also enriched at compensatory , gained , and strongly gained ( human-specific ) occupancy sites , but not at lost ( mouse-specific ) occupancy sites ( Fig . 4C–F , p<10−5 for each comparison versus lost ) . Importantly , we observed that active regulatory states are more enriched for both conserved and strongly gained TF occupancy sites than for compensatory or all gained sites ( Mann-Whitney test , p<10−5 ) , but we did not observe a difference in enrichment between conserved and strongly gained sites ( p = 0 . 56 ) or between compensatory and gained sites ( p = 0 . 54 ) . This pattern of regulatory chromatin enrichment was replicated in K562 cells , which themselves appear to have a similar chromatin state to human ProEs , supporting the functionality of these definitions ( S10A–E , G Fig . ) . In mouse ProEs , conserved TF occupancy sites were also enriched at active chromatin states and mouse-specific occupancy sites , while no enrichment was observed at strong human-specific sites , suggesting first that conserved TF occupancy sites are functional and preserve strong regulatory chromatin structure across millions of years of evolution ( S10F Fig . ) . However , it also suggests that there is a dramatic change in chromatin structure at orthologous human- and mouse-specific TF occupancy sites ( 4F Fig . ) . To determine if functional changes in transcription are associated with alterations in TF occupancy during the course of evolution , we investigated occupancy near species-specific genes ( see materials and methods ) . We discovered that human-specific genes are significantly enriched for human-specific ( gained and strongly gained ) TF occupancy ( Fig . 4G ) . Corresponding to this observation , mouse-specific genes are significant enriched for mouse-specific ( lost ) TF occupancy sites . Surprisingly , these genes are also enriched for conserved and compensatory occupancy sites , a finding that we investigate more thoroughly below . Although the direction of causality is difficult to determine , we suggest that master TFs partially drive epigenomic evolution at orthologous genomic regions by mediating changes to the underlying chromatin structure . Indeed , it has been shown in corresponding null cell lines that the addition of master TFs , such as GATA1 , can remodel chromatin structure to increase transcription of certain genes , but our results suggest that master regulators play a far more important global role in chromatin remodeling during evolution [39] . Alternatively , de novo chromatin remodeling may impair the ability of TF complexes to bind , resulting in the transcriptional changes observed . We sought to understand the functional consequences of the observed epigenomic differences by quantifying the extent to which changes in chromatin structure and master regulator occupancy explain transcriptional divergence between species during terminal erythroid differentiation . We verified , using time-series RNA-seq data of gene expression , that intra-species transcription is indeed more conserved than inter-species transcription ( S11 Fig . ) [11] . For example , the gene expression profiles of late stage human OrthEs are more similar to early stage human ProEs than they are to mouse OrthEs ( S11 Fig . ) . We observe that the matching early progenitor states ( ProEs and BasoEs ) are more similar to their species-specific erythroid cell model , K562 or G1E , than to corresponding stages across species ( S11 Fig . ) . Intensity of epigenomic marks around TSSs has been shown to explain up to ∼50% of gene expression , providing a simple framework to globally investigate species-specific differences in transcription [40] . We derived a naïve predictive model of transcription in ProEs based upon total epigenomic mark intensity in promoter regions using linear regression with an L1 penalty . Our derived models of ProE gene expression learned across both species using six histone modifications and four TFs performed well: without over fitting , these models are able to explain between 58% and 61% of the variation in gene expression for each species based upon the coefficient of determination ( R2; Fig . 5A-B , D ) . Models learned independently on each species resulted in similar parameters and were unable to perform better , confirming that transcriptional “rules” are strongly conserved across species ( S6 Table ) . This model remained highly predictive throughout terminal erythroid differentiation , providing further evidence that most epigenetic modifications are dynamically determined at the ProE stage ( S6 Table ) . Interestingly , in this model , chromatin modifications and not TF-occupancy , were most predictive of gene expression ( H3K9ac , H3K4me3 , H3K27me3 , H3K36me3 , and GATA1 in order of importance , Fig . 5E ) . Having confirmed the biological significance of our model , we applied it to model differences in gene expression between species based upon changes in epigenetic marks . Utilizing this approach , we are able to explain 18% of the changes in gene expression between species . Considering that most genes are not differentially expressed between species , we applied our model to only species-specific expressed genes ( see materials and methods ) . In this case , we are able to explain 34% of the variation in gene expression between species based solely upon promoter epigenetic mark ( Fig . 5C ) . Although our promoter model was highly predictive and elucidated functional biological divergence , we further address the possibility that transcriptional changes are also associated with TF occupancy without restricting our analysis to only promoter regions . Specifically , we investigated the hypothesis that the evolutionary loss or gain of TF occupancy at CRMs is indicative of changes in nearby gene expression . We summarized time-series gene expression profiles for each of the categories of TF conservation that we defined previously ( conserved , gained and strongly gained ( human-specific ) , lost ( mouse-specific ) , compensatory , and two additional subcategories of conserved occupancy sites , Fig . 5F and S12 Fig . ) . Across all TFs , genes proximally occupied by at least one conserved TF were expressed at significantly higher levels across all cell states , from ProE to orthochromatic erythroblasts ( OrthEs ) ( Fig . 5F and S12 Fig . ) . The sequential ordering by differential gene expression of groups associated with different TF occupancy ( conserved > strongly gained > compensatory > gained > lost ) is identical to the ordering of these groups based upon their association with active regulatory states . Furthermore , limited evidence suggests that while gene expression is most similar between groups at terminal stages , loss or gain of master TF occupancy may affect the timing of gene expression , resulting in subtle differences in expression during differentiation ( S12 Fig . ) . Applying this method to cross species differences in expression , we discover that genes proximally occupied by a strongly gained TF site show human-specific expression during terminal erythroid differentiation ( Fig . 5G , S13 Fig . ) . Interestingly , the genes that show the strongest mouse-specific expression are , first , occupied by one or more TF sites that are conserved across species but , second , have one or more mouse-specific TF occupancy site . These findings suggest that the gains and losses of TF occupancy sites are associated with changes in transcription across species . Moreover , we remark on the observation that even though a single TF may be conserved across species , changes in the occupancy of other TFs at nearby regions may have large functional effects , similar to previously reported results [35] . As a general principle , our observations show that conserved TF occupancy across species is associated with both strong gene expression and active regulatory states . Indeed , while this principle has been shown for conserved GATA1 DNA binding motifs on a small scale , we have confirmed this principle for multiple master regulators with biochemical data across both genomes [41] . Slightly attenuated patterns are observed for species-specific TF occupancy , while orthologous genomic regions of lost TF occupancy show little enrichment for active regulatory states and are indicative of low gene expression . Furthermore , these data suggest that not all species-specific occupancy sites have an immediately observable function: only strong human-specific occupancy was clearly associated with actively transcribed genes . To illustrate specific features of epigenomic conservation and divergence during cellular differentiation , we examined a few well-known regulatory regions involved in erythropoiesis , leveraging our framework to gain further insight into the physiological relevance of these differences . We first describe two regions of general epigenomic conservation with subtle , but important , differences . We investigated the well studied locus control region ( LCR ) of the developmentally regulated β-like globin genes [42] . In both species , the LCR consists of 5 closely spaced regulatory regions , termed hypersensitive sites ( HSs ) , directly upstream of the embryonic and adult β-like globin genes ( Fig . 6 ) . Each region in the LCR has been shown to loop to developmental stage-specific β-like globin genes to promote high-level gene expression [43] , [44] . Here , we investigate the conservation of TF-occupancy and chromatin state assignment at the first four HSs . We observe that the TF-occupancy profiles of these HSs are strongly conserved across species ( Fig . 6A–B ) . In particular , GATA1 and TAL1 bind strongly to each HS . Canonical DNA motifs conserved across 46 vertebrates and present in both human and mouse genomes can be identified ( Fig . 6C ) . Interestingly , we observe that the 1st and 3rd HSs are two of the highly constrained CRMs co-occupied by GATA1 , TAL1 , and KLF1 , confirming that these HSs are under strict evolutionary constraint . We observe stronger KLF1 intensity at HS1 compared to HS2 in human , but stronger KLF1 intensity at HS2 versus HS1 in mouse . Additionally , we observe increased NFE2 occupancy at HS3 in humans compared to the same region in mouse . Furthermore , the first four HSs in the LCR are associated with strong/poised enhancer states in humans whereas they are associated with strong/weak enhancer states in mice . Finally , although GATA1 and TAL1 occupy HS4 across species , the specific binding sites in this HS appear to be different for human and mouse ( Fig . 6C ) . Overall , chromatin structure and TF occupancy at the β-globin LCR is largely conserved , but subtle differences may have effects on stage-specific transcriptional patterns . We next turned our focus to the large 2nd intron of BCL11A containing an erythroid specific enhancer that , when disrupted in mouse cell lines , reduces BCL11A transcription and has been suggested to underlie common genetic variation of this key globin switching factor [45] , [46] . This enhancer is occupied by GATA1 and TAL1 in humans and contains a GATA1/TAL1 motif that is partially disrupted by the minor allele of the common polymorphism , rs142707 ( degenerative TAL1 motif is CAT for the wildtype and CAG for the minor allele , S14A–C Fig . ) . Although this binding motif is conserved across species , the guanine minor allele in humans is the ancestral allele present in other primates and mice ( S14C Fig . ) . We observe that GATA1 and TAL1 are also enriched at this site in mouse ProEs , suggestive of a conserved function for this enhancer element ( S14A Fig . ) . Nevertheless , we more broadly observe divergent patterns of TF occupancy , histone modifications , and gene expression across species , suggestive of functional differences at this locus ( S14A–B , D Fig . ) . This finding emphasizes that caution must be applied when investigating and interpreting results from single TF occupancy or HS site data alone rather than a comprehensive approach that includes multiple factors and histone modifications across a broader region . Next , we focus on two examples that show substantial divergence across the epigenome . Recessive mutations in SEC23B have been implicated in congenital dyserythropoietic anemia type II ( CDA II ) , but an erythroid phenotype could not be recapitulated in mouse models [12]–[14] . One hypothesis for this observation is that while SEC23A is not expressed in similar human cell-types , Sec23a is expressed in mouse and is functionally able to compensate for the absence of Sec23b , resulting in the absence of a phenotype in Sec23b knockout mice [10] , [47] . We therefore investigated these potential differences in transcriptional regulation . We observed no clear differences in TF occupancy or histone state at SEC23B , and this gene is similarly expressed between species ( Fig . 7E , S15A–B Fig . ) . Thus , we focused on SEC23A . While we observed some small differences in TF occupancy between species , the most striking difference is that the local region surrounding human SEC23A is in a general state of heterochromatin ( state 13 ) or polycomb repression ( state 14 ) , whereas the region around mouse Sec23a is comparatively open for transcription ( Fig . 7G–H ) . Expanding out to a small region around SEC23A , three homologous genes are present in both species and exhibit similar species-specific chromatin states as well as similar gene expression pattern corresponding to their matching SEC23A/Sec23a gene ( Fig . 7F–H ) . This suggests that transcription in the local region around SEC23A is repressed in humans whereas the homologous region around mouse Sec23a is significantly more transcriptionally permissive . This finding not only provides evidence for why knockout of Sec23b in mouse does not recapitulate the human disease phenotype , but also highlights a principle of epigenomic divergence: in concordance with our simplified promoter model ( Fig . 5A–C ) , the local genomic region has transitioned during evolution from a low/active state in mouse to a repressed state in humans , and transcription has been blunted as a result . Alternatively , the reverse possibility may have occured: for an unknown reason transcription has decreased in this region , driving the chromatin changes that are observed . Finally , we investigated a locus of interest where a gain of TF occupancy in a non-homologous genomic region is associated with species-specific gene expression ( observed globally in Fig . 5F–G ) . Growth differentiation factor 15 ( GDF15 ) is one of the most highly expressed genes in human differentiating erythroblasts but is absent in mouse erythroblasts ( Fig . 7A ) [11] , [48] . GDF15 has been suggested to play an important role in the regulation of iron homeostasis as a result of changes in the extent and effectiveness of erythropoiesis [49] . Patients with β-thalassemia and other diseases characterized by ineffective erythropoiesis show increased levels of GDF15 expression [50] . By analyzing epigenomic patterns at this locus , we identified a species-specific difference in chromatin structure at the GDF15 locus: human GDF15 in ProEs has a strong promoter , whereas mouse Gdf15 in ProEs has a poised promoter ( Fig . 7B–C ) . Most importantly , while we identified some TF occupancy near mouse Gdf15 , we identified a novel , putative CRM occupied by GATA1 , TAL1 , KLF1 , and NFE2 upstream of human GDF15 that is absent from the larger region that encompasses mouse Gdf15 ( Fig . 7B–D ) . Comparing the underlying genomic sequence of this putative enhancer across species , we found that this region is highly conserved in primates , but not in mice ( Fig . 7D ) , suggesting that human GDF15 expression may be driven by this element that is absent from mouse .
While numerous studies have been performed to understand how epigenomic modifications play a role in mediating cellular differentiation , only a limited number of studies have examined how these modifications have been altered during the course of evolution [29] , [32] , [34] , [51]–[53] . Here , we have used erythropoiesis as a model of cellular differentiation to study how epigenomic modifications can underlie evolutionary changes in gene expression . We performed a systematic comparative analysis of occupancy for six histone modifications and four master TFs in both human and mouse primary ProEs as well as in erythroid cell lines , integrating our results with high quality gene expression data . Models based upon promoter marks were highly predictive of gene expression and were nearly identical for both human and mouse ProEs . While we observed that chromatin modifications , at least in promoter regions , were generally conserved across species , subtle differences in H3K9ac , H3K4me3 , H3K27me3 , and H3K9ac were associated with differential gene expression between species . This finding partially accounts for the previously reported divergence in gene expression during terminal erythroid differentiation across species [10] , [11] . We found that only ∼25% of GATA1 , TAL1 , and NFE2 occupancy sites present in mouse ProEs were conserved in human ProEs; however , the loss of these sites is often offset by the acquisition of nearby compensatory TF occupancy sites . To some degree , compensatory sites appear to buffer transcriptional changes that occur from the original loss . This finding is consistent with the reported conservation and compensatory action of master regulators in other cell types that are found between closely related species [33] , [34] , [53] . In juxtaposition to other master regulators , KLF1 occupancy was highly conserved between human and mouse , approaching the conservation rates of TFs in closely related species of insects [52] , [53] . Acting in combinatorial fashion with GATA1 and TAL1 , we show that CRMs co-occupied by these three TFs are under strong evolutionary constraint and localize to genes that play a key role in defining erythroid cell state . The critical role of these TFs in defining erythroid cell state is highlighted by human genetic studies that have identified causal mutations for various forms of anemia in GATA1 and KLF1 [54]–[58] . We suggest that disruption of these modules in either human or mouse progenitor cells would severely compromise terminal erythroid differentiation , and thus these regions may harbor non-coding polymorphisms in humans that underlie human erythroid disorders . In particular , polymorphisms that disrupt a GATA1 , TAL1 , or KLF1 binding motif in conserved or strong species-specific occupancy sites would be leading candidates for causal mutations in these disorders . For example , we identified that GATA1 and TAL1 co-occupy the first intron of UROS in both human and mouse . While coding mutations in UROS have been identified in over 50% of patients with congenital erythropoietic porphyria , rare mutations that disrupt a constrained GATA1 binding site in the first intron of UROS have been found in patients lacking a putative coding mutation [59] . Indeed , in an era when whole genomes of patients with various diseases can be readily sequenced , identification of causal mutations is frequently a difficult problem [60] , [61] , and results from this study could help prioritize variants identified by such approaches . Furthermore , in Mendelian diseases where a pathogenic coding variant is not immediately identifiable , targeted sequencing of conserved TF occupancy sites near causal genes could prove useful as an inexpensive and likely high-yield approach in comparison to whole genome sequencing . In contrast to mouse ProEs , we found that TF occupancy in K562 cells is strongly conserved with human ProEs . We suggest that for certain erythroid disorders , K562 cells may more faithfully recapitulate features of the disease than primary mouse cells , particularly in cases where epigenetic or transcriptional regulation may be disrupted . The framework we have created provides an opportunity to prospectively ascertain the extent of conservation between mice and humans for various aspects of the transcriptional landscape underlying erythroid differentiation . In this study , we confirmed and uncovered multiple principles of epigenomic conservation . We found that conserved TF occupancy between species is strongly associated with active regulatory regions and strong transcriptional activity during terminal erythroid differentiation . Similarly , the strongest human-specific TF occupancy sites were also associated with regions of active regulation and strong transcription . When extrapolating on information gained from TF occupancy in mouse ProEs , it is important to consider not only that ∼75% of regions are not conserved in humans , but also that regions of lost TF occupancy exhibit reduced regulatory modifications as well as , on average , reduced transcription of genes across all stages of terminal erythroid differentiation . As a result , we emphasize the importance of using such a comparative framework when examining whether findings from mouse models of erythropoiesis may have relevance to human blood production . We have used our framework to interrogate specific regulatory regions as well as genes important in erythropoiesis to illustrate and provide vignettes for the principles that we identified on a more global scale . In particular , the results we present provide evidence that human GDF15 is actively transcribed and contains unique CRMs not found near mouse Gdf15 , consistent with its reduced expression in mouse erythroid cells . This finding is important when interpreting the role of Gdf15 in mouse models , and further investigation on the epigenetic regulation of GDF15 may help explain variation in iron and erythroid homeostasis between mice and humans . In other cases the epigenomic landscape is more conserved , such as at the β-globin or BCL11A gene loci , although subtle variation may explain the species-divergent gene expression patterns that are observed . Some limitations should be considered when interpreting the results of our study . First , while we included over 50 ChIP-seq datasets in our analysis , there are other histone modifications ( e . g . H3K27ac ) , TFs ( e . g . ZFPM1 and SPI1/PU . 1 ) non-coding RNAs , and methylation patterns that may be important for understanding species-specific differences in erythropoiesis . Furthermore , we cannot exclusively rule out the possibility that certain peaks are “hyper-ChIPable” due to a lack of IgG control datasets , although recent work provides convincing evidence that this consideration , while critical in yeast , is far less of a concern in complex metazoans [62]–[64] . Finally , although the ProE stage is ideal to investigate epigenetic changes that occur to mediate alterations in the transcriptional landscape [19] , [20] , [26] , [27] , we did not investigate temporal changes in epigenomic marks during earlier or more terminal stages of differentiation where species-specific differences may be more pronounced [34] . We have made all of our results publically available as filetypes that are quickly loaded into standard genome browsers ( IGV and UCSC Genome Browser ) . These data could guide investigators in choosing appropriate model systems for studying blood diseases or other aspects of erythropoiesis as well as aid in the interpretation of their results . Overall , our comparative epigenomics approach has successfully explained a significant portion of the transcriptional divergence observed during erythroid differentiation in mice and humans .
Hg19 and mm10 were used throughout the entire analysis as reference genomes for all human and mouse cell types , respectively . Orthologous genes were defined using Ensembl mouse to human ortholog matching and were downloaded from BioMart; genes which matched one: many were excluded from all analyses , resulting in 15506 one: one orthologous genes used for analysis . A smaller subset of orthologous genes ( 6596 ) with well-defined canonical transcripts from RefSeq was used for all quantitative promoter analyses . To compare mm10 to hg19 , the UCSC liftOver tool was used to lift coordinates over from one genome to another with one: one matching and 10% sequence conservation required . PAVIS was also used to annotate genomic regions such as TF-occupancy peaks based upon proximity to known genes [65] . ChIP-seq datasets were either downloaded from NCBI GEO or from the ENCODE project's homepage ( S1 Fig . ) . SRA files were transformed to FASTQ using FASTQ -dump from the NCBI SRA toolkit ( https://www . ncbi . nlm . nih . gov/books/NBK158900/ ) . Raw reads were aligned to the hg19 and mm10 genomes using Bowtie v0 . 12 . 9 with options “-v 2 -m 3 —strata –best” [66] . The BEDTools suite was used for multiple operations , comparisons , and intersections of all resultant BED files [67] . Reads were extended to a fragment length of 200bps , normalized to million-mapped-reads , and control input ( in million-mapped-reads ) were subtracted . In all quantitative analyses , reads were log2 scaled and read into R 3 . 0 . NGSplot was used to plot normalized average intensity curves across 15506 orthologous genes ( -2000 from TSS to +2000 after TES ) for all ChIP-seq datasets [68] . TF-occupancy peaks were initially called using MACS 1 . 4 to estimate fragment size [69] . When replicates were present ( e . g . GATA1 and TAL1 , S1 Fig . ) , MM-ChIP was used to combine and robustly call peaks from datasets across multiple laboratories and technical conditions to create sets of high-quality peaks [70] . The top n-percentile of each set of peaks was defined based upon the total number of mapped reads present in the peak region . When regions/peaks were lifted over from one species to another , the denominator used was always the number of regions which mapped to the new genome successfully from the original genome , while the numerator was the number of mapped regions that overlapped with the target region in the new genome . If peaks from a single TF were lifted across genomes , only a narrow region ( +/− 50 bps around summit ) was mapped to reduce the probability of incorrect mappings due to non-functional decrease in sequence similarity at the far edges of peak regions called by MACS . DNA motif enrichment was performed using MEME-ChIP in the MEME Suite with standard options [71] . E-values are reported as corrected p-values in all figures . Enrichment of combinatorial TF occupancy was assessed using 100 , 000 permutations across the genome with the Genomic Association Tester [72] . Chromatin states were estimated for 200 bp bins spanning both genomes using a Hidden Markov Model ( ChromHMM ) [38] , [73] . We settled on a 15 state model learned on all three cell types together , although we examined models ranging from 12 to 20 states ( Fig . 4A and S7 Fig . ) . Biological relevance for each state was assigned based upon frequency of chromatin marks and functional enrichments similar to previous studies [73] . For example , the “Active Transcription” state is marked exclusively by H3K36me3 and is enriched primarily for exonic regions whereas the “Strong Promoter” state is marked by H3K4me2 and H3K4me3 but not H3K4me1 and is enriched for TSSs . The final model was highly conserved between models derived exclusively from mouse ProEs , human ProEs , and K562 cells ( Fig . 4A , S8 Fig . , S9 Fig . ) . Enrichment of chromatin states across regions of TF occupancy was compared using the “OverlapEnrichment” command in ChromHMM . We performed multiple analyses to address the possibility that ChIP protocol differences may underlie the TF occupancy differences observed in analyses such as promoter differences and peak calling . First , we note that we were able to recover large sets of peaks ( >5000 peaks ) for each TF in each species , suggesting that immunoprecipitation was nominally successful . Indeed , western blots in human and mouse cell types suggest that these antibodies are specific for both human and mouse TFs ( http://genome . cse . ucsc . edu/cgi-bin/hgEncodeVocab ? ra=encode%2Fcv . ra&term=%22GATA1_ ( SC-266 ) %22 and 2TAL1_ ( SC-12984 ) %22 ) . Importantly , these peaks were all significantly enriched for the TF canonical motif ( Fig . 2D ) . This result is not surprising , given that human and mouse GATA1 and TAL1 share >86% similarity in amino acids based upon the Ensembl database . Furthermore , when we investigated occupancy site conservation between species by performing a sensitivity analysis on only the strongest 25% of peaks , the conservation rate of the least conserved TF ( TAL1 ) showed a similar increase to that of the most conserved TF ( KLF1 ) , suggesting limited bias between species ( S5 Fig . ) . Based upon this evidence , we believe that the most likely reason for the observed difference in absolute occupancy between species is that certain aspects of the protocols used vary between species and introduce bias such that weaker peaks in mouse ProEs may not be as readily observed in these cases . Alternatively , the difference in the number of occupancy sites could be a true biological observation . Regardless of the case , our estimates for conservation of TF occupancy scale with the absolute number of peaks called in mouse . In other analyses , we normalized between human and mouse to account for these differences . Single-end RNA-seq datasets of primary erythroblasts were downloaded from NCBI GEO ( GSE53983 ) [11] . K562 and G1E RNA-seq data was publically available from ENCODE and was also downloaded NCBI GEO ( GSE40522 ) and from the ENCODE website ( http://genome . ucsc . edu/ENCODE/ ) . Five human time points ( ProE , early BasoE , late BasoE , PolyE , OrthoE ) and four mouse time points ( ProE , BasoE , PolyE , OrthoE ) with three replicates each were used in this analysis . Similar to ChIP-seq data processing , SRA files were transformed to FASTQ using FASTQ-dump . RNA-seq data was processed with the Tuxedo Tools Suite using the same options as recent protocols , except CuffQuant and CuffNorm were used to derive normalized ( FPKM; fragments per kilobase of transcript per million mapped reads ) and raw count data for each transcript [74] . In particular , raw reads were aligned to the genome using TopHat v2 . 0 . 10 . Cufflinks v2 . 1 . 0 was used to assemble transcripts , CuffMerge was used to merge all annotations ( separated by species ) , and CuffQuant and CuffNorm were used to output data at the gene , TSS , and promoter level that was then imported to R . All Tuxedo Suite tools were run using standard options as indicated [74] . For predictive analyses of ProE gene expression , we first quantile normalized our epigenomic profiles to account for any species-specific biases in intensity and integrated this dataset with stage-matched RNA-seq data . In order to derive predictive models of transcription for each species without over fitting parameters , we performed standard linear regression with L1-penalization ( i . e . lasso regression , [75] ) . In this model , the β value of each predictor is shrunk towards zero until an optimum solution is reached; variables that add little predictive value are excluded ( β = 0 ) . Subsequently , 100-fold cross validation is performed for different lambda “penalization” values and a lambda one standard error lower than the best model was chosen to prevent over-fitting . In R , glmnet was used to perform L1-penalized linear regression [75] . We used the Integrative Genomics Viewer ( IGV ) to view epigenomic mark intensity files [76] . Bed files of aligned reads were extended to a fragment size of 200 bps and intensity files ( bigwig ) were created using UCSC Genome Browser tools . Enrichments were shown on a log scale unless otherwise noted and a cut-off of 20 bps was used as a lower bound in all representations unless otherwise noted . Phastcons nucleotide substitution rate score ( 0 to 1 ) and the primate and mouse ( mm10 ) to human ( hg19 ) alignments from the multiz 46-vertebrate multi-alignment were also used as available for import from the IGV servers [77] , [78] . We have made aligned and processed ChIP-seq data for all six histone modifications , four transcription factors , and derived chromatin states for all cell-types available on GEO at GSE59801 . Gene expression data processed in our pipeline is also available as aligned reads and as FPKM and counts for each gene . Furthermore , robustly defined transcription factor occupancy peaks and information regarding conservation , gain , or loss across species as discussed in the analyses is also made available in the same location . | The process whereby blood progenitor cells differentiate into red blood cells , known as erythropoiesis , is very similar between mice and humans . Yet , while studies of this process in mouse have substantially improved our knowledge of human erythropoiesis , recent work has shown a significant divergence in global gene expression across species , suggesting that extrapolation from mouse models to human is not always straightforward . In order to better understand these differences , we have performed a comparative epigenomic analysis of six histone modifications and four master transcription factors . By globally comparing chromatin structure across primary cells and model cell lines in both species , we discovered that while chromatin structure is well conserved at orthologous promoters , subtle changes are predictive of species-specific gene expression . Furthermore , we discovered that the genomic localizations of master transcription factors are poorly conserved , and species-specific losses or gains are associated with changes to the underlying chromatin structure and concomitant gene expression . By using our comparative epigenomics framework , we identified a putative human-specific cis-regulatory module that drives expression of human , but not mouse , GDF15 , a gene implicated in iron homeostasis . Our results provide a resource to aid researchers in interpreting genetic and epigenetic differences between species . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"medicine",
"and",
"health",
"sciences",
"genomics",
"hematopoiesis",
"gene",
"regulatory",
"networks",
"molecular",
"development",
"gene",
"expression",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"anemia",
"hematology",
"cell",
"differentiation"
] | 2014 | Altered Chromatin Occupancy of Master Regulators Underlies Evolutionary Divergence in the Transcriptional Landscape of Erythroid Differentiation |
Active sensing organisms , such as bats , dolphins , and weakly electric fish , generate a 3-D space for active sensation by emitting self-generated energy into the environment . For a weakly electric fish , we demonstrate that the electrosensory space for prey detection has an unusual , omnidirectional shape . We compare this sensory volume with the animal's motor volume—the volume swept out by the body over selected time intervals and over the time it takes to come to a stop from typical hunting velocities . We find that the motor volume has a similar omnidirectional shape , which can be attributed to the fish's backward-swimming capabilities and body dynamics . We assessed the electrosensory space for prey detection by analyzing simulated changes in spiking activity of primary electrosensory afferents during empirically measured and synthetic prey capture trials . The animal's motor volume was reconstructed from video recordings of body motion during prey capture behavior . Our results suggest that in weakly electric fish , there is a close connection between the shape of the sensory and motor volumes . We consider three general spatial relationships between 3-D sensory and motor volumes in active and passive-sensing animals , and we examine hypotheses about these relationships in the context of the volumes we quantify for weakly electric fish . We propose that the ratio of the sensory volume to the motor volume provides insight into behavioral control strategies across all animals .
The motor volume is the swept volume of positions a body occupies for a given trajectory or set of trajectories . It is a function of the way the body moves , as well as the geometric extent of the body . To define it more clearly , we first present a definition of the time-limited reachable set . We then use this definition to informally define the motor volume and the stopping motor volume ( the swept volume of the body over trajectories that bring the body to a halt ) , with the precise definitions presented in Text S1 . For simplicity of description , we treat an organism as a rigid 3-D body . We define the configuration space as the six-dimensional space representing the rigid-body degrees of freedom ( typically the ( x , y , z ) position of the center of mass , and θ , ϕ , and Ω , the yaw , pitch , and roll angles , respectively ) . The state space X consists of the six configuration components and their time rates of change ( vx , vy , vz , vθ , vϕ , vΩ ) , resulting in a total of twelve dimensions . The dynamics of the system are given by where x is the instantaneous state and u is the instantaneous control input from the space U℘ of all feasible instantaneous control inputs . The time-limited reachable set R ( x0 , T ) is a construct from nonlinear control system theory [14–16] referring to all points in the state space X that can be reached by a system of the form given by Equation 1 from an initial state x0 given any feasible control history u of duration T . A feasible control history is a control input to the system as a function of time , such as muscle activations for a musculoskeletal system or steering wheel angles for a car . The reachable set is defined as: This is the set of states reachable in time exactly T . In our subsequent definitions , we will use the union of all reachable sets from t = 0 to t = T , denoted as: To illustrate the concept of the time-limited reachable set , consider the simple one-dimensional case of a locomotive moving along a train track ( Figure 4A ) . The state space X is simply the locomotive's configuration space ( the x position of its center of mass ) , as well as the time rate of change of this one configuration component ( vx ) . The initial condition x0= ( 0 m , 5 m/s ) . The control inputs are limited to the set U℘ = [–1 m/s2 , +1 m/s2] . Figure 4A shows two dashed curves representing state space ( x , vx ) trajectories under constant acceleration ( +1 m/s2 ) and deceleration ( −1 m/s2 ) , starting from an initial velocity of 5 m/s and running for 5 s; all time-limited reachable sets of 5 s or less from this initial condition must be within these two curves . For example , for t = 3 s , the time-limited reachable set R [ ( 0 m , 5 m/s ) , ≤ 3 s] is the union of sectors “1” through “3 . ” Figure 4B shows time-limited reachable sets R [ ( 0 m , 0 m/s ) , ≤ 5 s] —identical conditions as for Figure 4A but with zero initial velocity . For organisms with internal degrees of freedom ( multi–rigid-body systems with joints or flexible bodies ) , the same concepts apply but now the state space needs to include the additional degrees of freedom , and their time rates of change . The motor volume ( Text S1 , Equation S1 ) is similar to the reachable set but with several important differences . First , rather than a set of points in state space , the motor volume is the volume defined by the set of ( x , y , z ) coordinates of all positions occupied by the body over the time interval of interest . Second , these points are in a coordinate frame that is aligned to the body at the onset of the behavior for which the motor volume is being constructed . For the locomotive example , the thick green line between x = −2 and x = 38 in Figure 4A represents the motor volume ( more correctly , the motor “line” in this case ) MV ( 5 m/s , ≤ 5 s ) . As defined in Text S1 , the motor volume is derived from the time-limited reachable set , where this set is computed from Equations 1 and 3 [17 , 18] . However , in our case , we do not have access to either the equation of motion ( Equation 1 ) for the fish or to the set of feasible control histories; thus , we will estimate the motor volume empirically by examining motion capture data collected from over 100 prey capture sequences ( see Methods ) . In this paper , we will consider the motor volume of the fish over all initial velocities that are typical within our prey capture sequence dataset . Prey capture sequences were selected to begin about 0 . 5 s before the onset of behavioral response and end at prey engulfment [11] , and thus contain both pre- and post-detection behavior . The stopping motor volume ( Text S1 , Equation S2 ) designates the set of all ( x , y , z ) coordinates of all positions occupied by the body from a given initial velocity to the location at which the body can be halted in the shortest amount of time . For the locomotive example , MVstop ( 5 m/s ) is equal to MV ( 5 m/s , ≤ 5 s ) with control inputs limited to the set U℘ = [–1 m/s2] , and is shown by the red line in Figure 4A . If the train had started at rest , then MVstop would collapse to a single point as illustrated by the red dot in Figure 4B . In this paper , MVstop will be ascertained through analysis of motion capture data ( see Methods ) ; if a mechanical model exists , MVstop can be estimated by computing optimal stopping trajectories from each initial velocity . The sensory volume ( SV ) for a given object is the volume defined by the set of ( x , y , z ) coordinates at which that object can be reliably detected , in body-fixed or sensory system–fixed coordinates . The SV depends on a number of factors , such as target size , orientation , velocity , properties of the sensory system , properties of the detection algorithm , and so on . In this paper , we will estimate SV ( prey ) for the active electrosensory detection of small water fleas ( D . magna ) . With these definitions in hand , we can explore possible functional relationships between sensory and motor volumes . Restricting our consideration to the stopping motor volume , consider Figure 5 , which shows three possible relationships between sensory and stopping volume geometries . In Figure 5A ( “collision mode” ) , the sensory volume is smaller than the stopping volume . We hypothesize that this situation should rarely occur in nature when the SV is for objects that the animal needs to avoid colliding with . Anywhere that MVstop is not covered by the SV represents a region in which unintended collisions could occur . Figure 5B ( “reactive mode” ) illustrates a situation in which the SV and MVstop are fully overlapping . We hypothesize that this condition represents a functional lower bound on the size and shape of the SV , when the SV is computed for obstacles that the animal should avoid . Also , if the animal's prey capture strategy demands that it come to full stop to consume or “handle” the prey , then the reactive mode can be considered a lower bound on the size and shape of the SV for prey capture . Finally , Figure 5C ( “deliberative mode” ) illustrates the case in which the SV is much larger than , and fully encloses MVstop . This case is typical for visually guided predators in terrestrial environments hunting in full daylight , where visual range can far exceed stopping distance . In the context of foraging behavior , the deliberative mode would generally lead to higher foraging efficiency . However , for active-sensing systems , the sensing range could be limited by factors such as energetic costs , clutter , and quartic attenuation , among others , leading to a situation closer to the reactive mode ( Figure 5B ) than the deliberative mode ( Figure 5C ) . If these constraining factors are particularly severe , then we predict that the SV should approximate MVstop , as in the reactive mode . Finally , we hypothesize that if the SV and MVstop approximately match , then the animal may make behavioral adjustments to either the sensory volume ( e . g . , by changing the probe intensity ) or to the stopping volume ( e . g . , by changing their “cruising” velocity ) to maintain a matched relationship between the SV and the MVstop , particularly where the absence of such adjustments would bring the animal into collision mode . Here we examine SV–MV relationships in the context of prey capture behavior of the black ghost knifefish . From earlier behavioral studies , we know that the fish comes to a near halt before engulfing prey [11] . Thus , we predict that SV ( prey ) should fully enclose MVstop over the velocities that characterize pre-detection swimming . Our results will show that the relationship between SV ( prey ) and MVstop for the black ghost is best described by the reactive mode shown in Figure 5B . For this mode , we predict that the animal might use behavioral adjustments to maintain a match between SV and MVstop . We test this hypothesis by reanalyzing results of our prior work with prey capture behavior in water of different electrical conductivities , which influences the distance at which prey can be detected [11] . The evidence suggests that as the SV ( prey ) shrinks in size due to increased water conductivity , the fish decreases its mean prey search velocity , which causes a corresponding decrease in the size of MVstop .
Using synthetic prey capture trajectories , the sensory volume for active electrosensory prey detection in A . albifrons was estimated by computing a detection isosurface surrounding the fish , such that every point on the surface generates a threshold level of activation after summing and filtering the electrosensory afferent signals ( see Methods ) . Every point contained within this bounding isosurface would yield a suprathreshold signal . The estimated SV was found to be omnidirectional ( Figure 3 ) , extending in all directions from the body surface . On average , the sensory surface was 34 ± 5 mm ( N = 7 , 056 detection points , all numbers quoted as mean ± standard deviation unless otherwise noted ) from the fish's body surface . As described in more detail below , this is also consistent with the empirically determined Daphnia detection distance of 28 ± 8 mm ( N = 38 ) reported in an earlier behavioral study [19] once the sensorimotor delay time is factored in . In addition to using synthetically generated linear prey capture trajectories , we also tested detection performance using empirically measured fish and prey trajectories from a previous study , which have more complex spatiotemporal profiles [11] . The computationally estimated detection distance was 33 ± 7 mm ( N = 38 trials × 10 repeats , or 380 , see Methods ) , compared with the measured detection distance of 28 ± 8 mm ( N = 38 trials ) . However , the latter empirical value is actually the “reactive distance” at which a motor response is first observed and does not include the sensorimotor delay between detection and movement . The estimated sensorimotor delay for behavioral reactions in the knifefish is 115 ms ( see Methods ) . Incorporating the sensorimotor delay , we obtain an estimate for the “neural” detection distance in the empirical data of 35 ± 9 mm ( N = 38 ) , which is in good agreement with the simulation result of 33 ± 7 mm for these trajectories . We examined the motor volume as defined above for the entire body , MVbody as well as the motor volume for the mouth alone , MVmouth , at fourteen discrete times , ranging from 117 to 700 ms . Motor volumes at three of these time points are illustrated in Figure 6 . The motor volume that maximally overlaps the SV ( t = 432 ms ) is shown in interactive 3-D Figure S4 along with the SV . Because of the unusually high maneuverability of the fish , including its backward-swimming capability , the shape of the body motor volume is omnidirectional and approximately cylindrical on short time scales , extending both in front of the head and behind the tail of the fish . The size and shape MVmouth , while also omnidirectional , is more compact in the rostrocaudal direction . Figure 7 shows the relationship between the SV and the stopping motor volume . The SV fully encloses MVstop except for a rostral protrusion .
In the present study , we find that the time-limited motor volumes for A . albifrons are omnidirectional and extend equally in front and behind the fish ( Figure 6; interactive 3-D version Figures S2 , S3 , and S5 ) . This confirms the previously reported propensity of weakly electric knifefish to spend a significant fraction of time swimming backward at velocities comparable to their forward-swimming velocities [13] . The motor volume is found to be cylindrical , indicating that the fish has considerable lateral and dorsoventral mobility [11 , 12] . The time-limited motor volume introduced in this study provides a quantitative measure of the fish's motor capabilities as a function of time interval . The body motor volume was closer in shape to the sensory volume and exhibited greater overlap than the mouth motor volume ( Figure 6 ) . Functionally , the mouth motor volume is more closely associated with prey capture behavior , whereas the body volume may be more relevant to obstacle avoidance and general navigation in complex environments . The sensory volume seems better matched to the body motor volume , which suggests that general navigational capabilities , habitat complexity , and obstacle avoidance should all be considered when examining relationships between sensory and motor volumes in the context of prey capture behavior . We find that the active sensory volume for prey detection is also omnidirectional and approximately cylindrical ( Figure 3 ) . The omnidirectional shape of the SV is similar to the shape of the isopotential surfaces of the self-generated electric field surrounding the fish ( Figure 1A ) , but exhibits less of a bulge in the caudal tail region due to anisotropies in the sensor density ( Figure 1B ) , field intensity ( Figure 1A ) , reduced sensory surface area due to the tapering body morphology ( Figure 1B ) , and the sensitivity of the primary electrosensory afferents to prey-induced perturbations [20 , 21] . The relative importance of these factors in determining the precise shape of the sensory volume was not examined . It will be particularly interesting in future studies to explore the extent to which the higher electroreceptor density in the head region of the fish influences prey-detection distance versus spatial resolution of prey position . The simulations of prey detection in this study combined quantitative models of each of these factors in order to arrive at an estimated sensory volume for electrosensory-mediated prey detection . As shown in Figure 2 , there was good agreement between the computationally estimated sensory volume and the empirical distribution of prey detections found in an earlier study [11] . The empirical study had relatively low N ( 38 prey capture events for the water conductivity that yielded maximum detection range ) , and the detection points were biased toward the dorsal surface of the fish , because prey were introduced into the tank from above . The computational approach used here allowed us to obtain a more complete and less biased estimate of the fish's sensory detection volume . Our prediction that the fish avoids collision mode is supported by Figure 7 ( interactive 3-D version Figure S6 ) , which shows nearly full enclosure of the stopping volume by the sensory volume . The stopping volume was taken from just before detection ( which always occurred during forward movement ) to the point of zero forward velocity as the fish reverses to capture the prey . Thus , unlike the time-limited motor volume , which sampled all initial velocities including negative velocities , the stopping volume is more strongly biased in the forward direction . However , the extent of the sensing range is restricted to nearly the lower limit of the reactive mode ( Figure 5B ) where the SV and MV are matched . We expect this is due to constraints that include the metabolic cost for emitting energy into the environment and the interference caused by interaction between the emitted signal and nearby clutter . Energetic costs scale steeply with sensing range , approximately as a quartic power of the range due to geometric spreading effects on the outbound and return paths of active probe signal [1 , 3] . For quartic scaling , doubling the active-sensing range would require a 16-fold increase in emitted energy . To appreciate the effect of this scaling , consider that an active 15-g A . albifrons requires approximately 300 J/day [22] . If we assume that ≈1% is used for the field , as estimated for another weakly electric fish [23] , then 3 J/day is needed for the field . To double the detection distance for D . magna from the measured ≈30 mm to ≈60 mm would therefore require 48 J/day for the field; to double this again to ≈120 mm ( still less than one body length for a 15-g fish ) would then require 768 J/day , more than double the entire energy budget of the fish . Thus , the high energetic costs associated with extending the active-sensing range is likely to place strong selective pressure on the shape and extent of the active sensory volume . In comparison , the high acuity passive visual system of a typical raptor allows it to spot prey from over a kilometer away , or about 10 , 000 body lengths . Although it is more difficult to provide a quantitative metric for the interference effects from clutter , the great reduction in emission power observed in dolphins and bats when surrounded by clutter or when nearing a target ( see Introduction ) suggests that this may also be an important factor in limiting the desirable range of an emitted signal . Returning to the automobile scenario , driving at night in a dense fog provides a practical example of where backscatter is a limiting constraint . Increasing headlight intensity under these conditions ( e . g . , switching from low beams to high beams ) can actually degrade detection performance because the “noise” from backscattered light makes it more difficult to detect the “signal” that is reflected back from a target of interest . Given that the black ghost is in reactive mode , we predict that the fish may make behavioral adjustments to either the sensory volume or to the stopping volume to maintain a matched relationship between the SV and the MVstop , particularly where the absence of such adjustments would bring the animal into collision mode . We are able to qualitatively evaluate this hypothesis by examining how search ( predetection ) swimming velocity varies with water conductivity . We have shown that water conductivity changes the range at which prey are detected [11] . In the previous study [11] , we found that the mean detection distance decreased from 28 mm at a water conductivity of 35 μS/cm to 12 mm at 600 μS/cm . Over this conductivity range , the fish's predetection swimming velocity decreased 30% from 99 mm/s to 71 mm/s . At the shorter detection distances associated with higher conductivity water ( 600 μS/cm ) , we have previously estimated that multiple sensory modalities , including passive electrosense and lateral line mechanosense , are playing a role [24] . Thus , quantitative evaluation of the matching hypothesis would require modeling the SV for these other sensory modalities , which is outside the scope of this study . The size of the estimated sensory volume , and to a lesser degree the shape of the volume , are influenced by the properties of the neural detection algorithm . A more sensitive algorithm will result in a larger detection range . The detection algorithm used here is not intended to model the fish's actual detection performance in detail . Doing so would require a more extensive analysis of additional factors , such as sensory reafference associated with tail bending , environmental background noise , contributions of other sensory modalities , neuroanatomical constraints on sensory convergence , etc . Rather , the detection volume reconstructed here is intended as an estimate of “best-case” detection performance based solely on changes in active electrosensory afferent spike activity . Echolocating bats emit ultrasonic energy into the environment to detect prey [25] . While the precise size and shape of the bat's sensory volume will vary with many factors ( species , call intensity , duration , etc . ) , the sensory volume for echolocation is generally a cone that extends in front of the head of the bat with an angular range of approximately ±30° in azimuth and elevation [26] . The angular coverage may extend as much as ±75° relative to the body axis when head and pinnae movements are included [25 , 27 , 28] . For the detection of flying insects by pipistrelle bats , the sensory volume extends at least 100–200 cm in front of the animal [25 , 29] based on the reactive distance to prey . The detailed shape of the bat's motor volume has not been reported . The stopping distance can be estimated by combining information on the initial velocity of the bat , maximal deceleration , and sensorimotor time delay . Taking a representative bat cruising velocity of 5 m/s [25] , a maximal deceleration of 15 m/s2 ( estimated from a sample trajectory in [25] ) , and an estimated sensorimotor delay of 100–200 ms [30] yields an estimated stopping distance in the range of 130–180 cm . Although there is a great deal of uncertainty in these estimates , the stopping distance of the bat seems comparable to the sensory range for prey detection . This suggests that bats , like electric fish , have an active sensory volume for prey detection that may be comparable to their stopping volume . Quantitative comparisons of sensory and motor volumes for a single bat species would help clarify these relationships . Odontocetes ( toothed whales , dolphins , and porpoises ) also use ultrasonic energy for prey detection . Dolphins can detect prey-sized objects at distances on the order of 100 m [6 , 31] . The 100-m sensing range of dolphins is certain to be significantly beyond their stopping volume , although there is little published data with which to make quantitative comparisons . This suggests that energetic and clutter-related constraints on active sensing may not be as significant for dolphins as they are for bats and electric fish . Both the active electrosensory volume for prey detection and the time-limited motor volume of A . albifrons are omnidirectional and approximately cylindrical . This is in striking contrast to most other animals , which tend to exhibit a strong forward bias in both sensory and motor volumes . This forward bias is observed for passive-sensing systems such as visually guided fish [32–37] . Figure 8 ( interactive 3-D version Figure S7 ) compares the omnidirectional prey sensory volume in the black ghost knifefish ( A . albifrons ) with a more typical forward-biased passive sensory volume . The latter is illustrated by the volume for visually mediated prey detection in the stone moroko ( Pseudorasbora parva ) , a fish of comparable size that also feeds on Daphnia [36] . The angular coverage of the visual volume is approximately 100° in azimuth and 60° in elevation , with a range that varies from about 60–120 mm , depending on environmental conditions [36]; the 120-mm range is shown in Figure 8 . The active electrosensory volume of A . albifrons and the passive visual volume of P . parva for prey detection are similar in size ( approximately 1 , 000 cm3 each ) but quite different in shape . We propose that the ratio of the sensory volume to stopping volume ( SV:MVstop ) is a useful metric for both active and passive sensory systems when considering whether sensorimotor control systems are in collision , reactive , or deliberative mode . Collision mode occurs when the ratio is below unity . Reactive mode occurs when the ratio approximates unity , as appears to be the case for knifefish and bats . In this mode , sensorimotor control algorithms are likely to be reactive , with relatively fast , direct coupling between sensation and action . Movement options are largely conditioned by mechanical considerations such as inertia and minimal turning radius . For example , some sensor-based motion planning algorithms in robotics are based on estimating the stopping volume for the nearest obstacle; as the robot becomes more massive , the range of any active-sensing system for obstacle detection must be extended accordingly [38] . Deliberative mode occurs when the ratio is large , as for dolphin echolocation and for many passive visual and auditory systems . In this mode , an animal can acquire sensory data from targets that are far outside its stopping volume . This allows the animal a greater range of movement options , because there is adequate time for complex motion planning before reaching the target [39] . For example , in the context of prey capture behavior , a dolphin with a high SV:MVstop ratio is able to engage in long-range tracking of distant prey , whereas a weakly electric fish with a ratio near unity must use a reactive strategy for chance encounters with nearby prey . Quantitative analyses of sensory and motor volumes for both active and passive-sensing systems can highlight important functional relationships between sensing , movement , and behavioral control in animals .
In a previously published study [11] , adult weakly electric fish ( A . albifrons ) were videotaped in a light-tight enclosure under infrared illumination . Individual water fleas ( D . magna , 2–3 mm in length ) were introduced near the water surface and drifted downward; prey capture behavior was recorded using a pair of video cameras oriented along orthogonal axes . Relative to the fish's velocity ( ∼100 mm/s ) the prey were relatively stationary ( prey velocity < 20 mm/s ) . Prey capture events ( from shortly before detection to capture ) were subsequently digitized , and 3-D motion trajectories of the fish surface and prey were obtained using a model-based tracking system with a spatial resolution of 0 . 5 mm and a temporal resolution of 1/60 s [40] . The time of prey detection was defined by the onset of an abrupt longitudinal deceleration as the fish reversed swimming direction to capture the prey [11] . These reversals are characteristic of most prey capture encounters . This is related to the fact that the fish tends to swim forward with its head pitched downward , such that the dorsum forms the leading edge as the fish moves through the water . Initial prey encounters thus tend to be uniformly distributed along the entire length of the body , so a reversal of swimming direction is typically required to intercept the prey . The volume of space supporting prey detection by the active electric sense was estimated computationally using measurements and empirically constrained models of the prey , electric field , fish body and sensor distribution , electrosensory images , afferent firing dynamics , and behavior . Model parameter values and their sources are summarized in Table 1 . Electric field . The electric field vector Efish ( mV/cm ) at a 3-D point in space x was computed using an empirically tested model of the electric field [41] . This model sums the individual contributions to the field from each of a series of charged poles used to model the electric organ of the fish: where x is a point in space ( cm ) , is the location of pole i of n total poles , q is a normalization constant ( mV cm ) that scales the overall magnitude of the field , σmes is the conductivity of the water that the field measurements were performed in ( which establishes the q value ) , and σmod is the conductivity of the water for the simulated field . The quantity q is analogous to electric charge in an electrostatic model and is distributed such that the first m poles have a “charge” of q/m and the remaining poles have a charge of –q/ ( n – m ) , resulting in a total net charge of zero . For our simulations , n = 267 , m = 266 ( all but one pole at the tail was positive ) , q = 10 , and the pole locations xp ran from the nose to the tail of the fish along the central axis of the fish body with equidistant spacing . These values resulted in field values within 10% of measurements of the electric field vector Efish of A . albifrons obtained by other researchers ( B . Rasnow , C . Assad , P . Stoddard , unpublished data ) in water of conductivity σmes = 210 μS/cm using a multiaxis electrode array [20 , 42] ( Figure 1A ) . The term scales the field strength to the water conductivity used in simulation σmod = 35 μS/cm . This scaling is based on empirical measurements of the knifefish field at different water conductivities [43] , which suggest the electric organ can be idealized as constant current source . We selected 35 μS/cm because our earlier study [11] found that the detection range was highest for trials at this conductivity , and this conductivity is most similar to rivers of the fish's native habitat . Body model with electroreceptor distribution . We used a prior survey [44] of the density of probability type ( P-type ) tuberous electroreceptor organs ( hereafter electroreceptors ) on the surface of A . albifrons . These are the dominant electroreceptor type for A . albifrons [45 , 46] . Each electroreceptor is connected uniquely to a primary afferent , which generates action potentials with a probability that varies with stimulus intensity . The receptor locations were mapped onto a high-resolution polygonal model of the fish derived from a 3-D scan of a body cast [19 , 24] in accordance with the measured sensor densities [44] ( Figure 1B ) . This resulted in a total of 13 , 857 mapped electroreceptors , in close agreement with neuroanatomically derived counts from A . albifrons [44] . Prey model . Based on prior measurements of live prey ( D . magna ) , it was modeled as a 1 . 5-mm-radius conductive sphere with an electrical conductance of σobj = 300 μS/cm [19 , 24] . Idealizing the prey as a sphere allows us to use an analytical model for the stimulus caused by the prey , described below . Electrosensory image formation . The voltage perturbation Δϕ ( mV ) at an electroreceptor on the fish surface , arising from a small spherical target object , was computed using an empirically tested model [47]: where Efish ( mV/cm ) is the electric field vector at the prey , r ( cm ) is the vector from the center of the spherical object to the electroreceptor on the fish surface , a is the radius of the sphere ( cm ) , σobj is the conductivity of the sphere , and σw is the conductivity of the water ( μS/cm ) . Simulations were run with water conductivity σw set to 35 μS/cm ( see Methods , Electric field ) . Primary afferent spiking activity . Computed voltage perturbations at each sensory receptor on the fish body were transformed into primary afferent spiking activity using a previously published adaptive threshold model of P-type ( probability coding ) primary electrosensory afferent response dynamics [48] . This model gives rise to negative correlations in the interspike interval ( ISI ) sequence , which lead to long-term spike train regularization . This correlation structure has been shown to increase information transfer and improve detection performance for weak signals [48–50] . The electric organ discharge ( EOD ) frequency was taken as 1 kHz , which is typical of A . albifrons [51] . P-type afferents fire at most one spike per EOD cycle . Thus , afferent activity was modeled as a binary spike train with a sampling period equal to the EOD period ( 1 ms ) . On each EOD cycle ( n ) , the following update rules are evaluated in order: Equation 6 implements low-pass filtering of the voltage perturbation Δϕ[n] with gain β and time constant τm . The state variable u[n] can be conceptualized as a membrane potential; it is initialized to 0 , corresponding to the steady-state value with no stimulus present ( Δϕ = 0 ) . Equation 7 adds random noise to u[n] to create a noisy membrane potential v[n]; the noise w[n] is modeled as zero-mean Gaussian noise with variance σ2 . The actual noise distribution is likely to be more complex , but the Gaussian approximation adequately captures available empirical data . Equation 8 describes the behavior of an adaptive spike threshold θ[n] that decays toward a baseline threshold θ0 with a time constant τθ . Equation 9 represents the process of spike generation , where s[n] is the binary spike output ( s = 1 , spike; s = 0 , no spike ) ; H is the Heaviside function , defined as H ( x ) = 0 for x < 0 and H ( x ) = 1 for x ≥ 0 . A spike is generated whenever the noisy membrane potential v[n] exceeds the threshold θ[n] . Equation 10 implements a relative refractory period by elevating the threshold θ[n] by an amount b immediately following a spike . ( The threshold level subsequently decays toward its steady state value according to Equation 8 . ) Parameter values common among all afferents in the model were β = 2 . 0 , b = 0 . 09 , σ = 0 . 04 , θ0 = −1 , and τm = 2 [48] . The values of the time constant τθ were generated independently for each afferent according to τθ = 21 – 18 ln ( z ) , where z is a uniformly-distributed random number between 0 and 1 . This distribution for τθ results in a distribution of baseline firing probabilities in the model that is matched to the empirically observed distribution [52] ( Figure 9A ) . The initial value for the threshold level θ[n] was drawn randomly from a Gaussian distribution with mean ( 0 . 064 ) and standard deviation ( 0 . 045 ) matched to the steady-state distribution of threshold values obtained from the model with no stimulus present ( Δϕ = 0 ) . As detailed in Brandman and Nelson [48] , these parameter choices result in a distribution of baseline firing rates , gains , and amplitude modulation–frequency tuning properties similar to empirically measured values [21 , 53] , as shown in Figure 9 . At the time of prey detection , the peak change in transdermal voltage is on the order of 1 μV , and the peak change in afferent firing rate is on the order of 1 spike/s [12 , 54] . Simulated prey encounter trajectories . During the search phase of prey capture behavior observed in earlier studies , A . albifrons typically swims forward with a mean longitudinal velocity of ∼100 mm/s , with its head pitched downward at an angle of ∼30° relative to horizontal [11] . The slowly moving prey were relatively stationary , with typical velocities less than ∼20 mm/s [11] . In the current simulation study , we modeled this relative motion between the fish and prey by moving the prey target along horizontal rays at 100 mm/s toward a stationary model fish pitched downward at 30° ( Figure 10A ) . Two sets of such prey trajectories were simulated , one set consisting of trajectories from head-to-tail , the other set from tail-to-head , since the fish swims forward and backward . A static model of the fish ( 140 mm long ) was centered within a rectangular box at a pitch of 30° . The box size was chosen such that any prey trajectory originating on a box face would begin well outside the detection range of the fish . The shortest distance between a point on any face of the box and any point on the fish was 60 mm . This distance was set by examining preliminary simulations and empirical measurements , which showed typical detection distances of 20–40 mm . The resulting dimension of the box was 245 mm ( length ) , 129 mm ( width ) , and 193 mm ( height ) . For the head-to-tail trajectories , horizontal prey positions from the front plane to the back plane were generated with starting and ending points centered on a grid with spacing of 5 mm ( a total of 1 , 014 trajectories ) . Variation in starting and ending points was achieved by adding random values ranging from negative to positive one-half the grid spacing . Intervening points on the trajectory were on a straight line , at a time interval of 1/60th of a second . The same number of tail-to-head trajectories were generated in the same manner . For each simulated trajectory , we use a detection algorithm to estimate the point at which the prey should have become detectable based on changes in afferent spike activity . The detection points that emerge from this analysis are then used to estimate the size and shape of the electrosensory prey detection volume . The approach used here is not intended to model the fish's actual detection performance in detail . Doing so would require a more extensive analysis of additional factors , such as sensory reafference associated with tail bending , environmental background noise , contributions of other sensory modalities , neuroanatomical constraints on sensory convergence , etc . Rather , the detection volume reconstructed here is intended as an estimate of “best-case” detection performance , based solely on changes in electrosensory afferent spike activity . The voltage perturbation corresponding to each prey position was computed from Equation 5 across the full complement of 13 , 857 sensory receptors . The resulting history of voltage perturbations at each receptor was interpolated to produce values at each millisecond and used as input for Equation 6 . Because of the stochastic nature of the afferent model , ten spike trains were generated for each afferent voltage history for each trial . Output spike trains for these synthetic trials were individually filtered using a boxcar filter with a window size τ = 200 ms , which in previous studies has corresponded to the best weak signal detection capability [52] . At the end of each millisecond , we extract and sum the scalar activity level of each of the 13 , 857 afferents . The motivation for this approach is that we assume that any neural detection algorithm will require pooling of information from multiple afferents; our goal here is not to specify exactly how this pooling process takes place , but rather to evaluate a best-case scenario . Using the “no-stimulus” condition , a detection threshold was established at a level that yielded one false detection per ten repeats . This detection strategy is a population-level extension of the algorithm described by Goense and Ratnam [54] for detection of weak signals in an individual spike train with ISI correlations . The prey detection distance was defined by the position of the prey at the time of threshold crossing for the “stimulus” condition . For each prey trajectory , the median and standard deviation of the detection distances were computed over ten repeats . The number of detections over all trajectories formed a bimodal distribution , with one peak near one , corresponding to “false alarms” for trajectories outside of the detection range , and a second peak near ten , corresponding to “true hits” for detections well within sensory range . There was a minimum in the bimodal distribution around seven , and we retained all trajectories with eight or more detections ( out of ten ) for further analysis . The mean and standard deviation of the detection distance for the resulting cloud of detection points was computed . To create the SV , this point cloud was triangulated into a smooth 3-D surface representation using a commercial software package ( RapidForm , INUS Technology , Seoul , South Korea ) . Measured prey encounter trajectories . The synthetic prey capture distance results were validated by following the same methodology outlined above using measured prey encounter trajectories from an earlier study [11] . The earlier study found best detection performance for trials in low conductivity water most similar to the water of the fish's native habitat . Therefore , we examined only this subset ( 35 μS/cm , N = 38 ) . In our prior study , the measured detection locations were estimated by examining the first change in behavior near to the prey [11] and therefore include sensorimotor delay time . Thus , to compare these to the computed neural detection points , we retrieved the distance between the prey and the fish at the detection time minus the sensorimotor delay time ( 115 ms [55] ) . The estimation of MV from motion capture data does not assume the fish is stationary at the initial time step; rather , MV is estimated over all the initial velocities typical of our prey capture behavioral segments starting from just before detection to capture , which includes forward as well as backward velocities , in addition to heave ( up in the body frame ) , and angular velocities such as roll and pitch . It is defined in the coordinate frame at the fish's initial position ( see Text S1 for details ) . We consider both the , where is the initial velocity in the coordinate frame fixed to the fish's initial position , for the entire body surface , MVbody , as well as the 3-D motor volume for the mouth alone , MVmouth . The MV was computed from the 3-D fish surface trajectory data obtained from 116 reconstructed prey-capture trials from an earlier study [11] . Because the motor volume for the mouth is a subset of this space , we will simply refer to the full volume as the MV , unless the distinction between mouth and body volumes is relevant . was estimated from the trajectory data at fourteen different times T , ranging from 117 to 700 ms . For a given T , the nodes on the surface of the polygonal fish model at time tinit + T were transformed back into the body-centered coordinate system of the fish at time tinit . This was repeated over all possible starting times tinit for each trajectory ( every 1/60th of a second up to the length of the trial minus T ) , thus uniformly sampling all observed velocities . The result of this procedure was a dense point cloud , representing where points on the fish's surface can reach over time T . The points on the surface of this cloud delineate an empirical estimation of ( Equation S1 ) in the absence of the equation of motion ( Equation 1 ) and feasible control histories , as discussed in the Introduction . The surface of the motor volume at each of the 14 intervals was defined by binning of the point cloud around the fish into voxels ( 5 × 5 × 5 mm ) , smoothing the data with a 3-D Gaussian convolution kernel ( standard deviation , 5 mm ) , and setting a threshold to include all voxels with point counts up to the 95th percentile . Each resulting point cloud was triangulated to form closed polyhedra for further analysis using commercial software ( RapidForm , INUS Technology , Seoul , South Korea ) . The motor volume for the mouth was constructed following the same procedure as for the body , but using only a single node at the rostrum of the polygonal body model . Because only a single node was used , the resulting point cloud was less dense than the body point cloud . To accommodate the lower density , we maintained all voxels with point counts up to the 90th percentile ( rather than the 95th percentile used for the body ) when constructing MVmouth . The stopping volume was constructed similarly , but for comparison of the stopping volume to the SV , we restrict our selection of trials to those of the same conductivity as used for estimating the SV , a total of 38 trials at 35 μS/cm . Unlike the MV , MVstop is not a function of a fixed time T but rather depends on the set of initial velocities from which we monitor movement until zero velocity is reached ( see Text S1 ) . Thus , we examine the volume swept by the body from the behavioral reaction ( detection ) time minus the sensorimotor delay time ( 115 ms , [55] ) to the time at which the longitudinal velocity of the body is zero . We take the union of these 38 volumes to derive the stopping volume over all 35 μS/cm trials . Computations were performed on a 54 CPU ( 2 GHz G5 , 1 GB RAM ) cluster of Xserves ( Apple Computer , Cupertino , California , USA ) running OS X . An open-source distributed computing engine ( Grid Engine , Sun Microsystems , Santa Clara , California , USA ) was used to manage the computation across the nodes . Simulation and analysis programs were written in MATLAB ( The Mathworks , Nantick , Massachusetts , USA ) and compiled to portable executables for execution on the cluster .
Interactive 3D visualizations of the sensory volume for prey ( water fleas , D . magna ) ( SV ) , time-limited motor volume ( MV ) , and stopping motor volume ( MVstop ) for A . albifrons , the black ghost knifefish . These Virtual Reality Markup Language ( VRML ) models can be viewed using downloadable web browser plugins and external viewers available for many platforms . As of the date of publication , one of the following is recommended , in order of preference: Cortona VRML plugin ( Windows and Mac OS X; available at: http://www . parallelgraphics . com/products/cortona ) . Octaga VRML player ( Windows , Mac OS X . and Linux; available at: http://www . octaga . com/ ) . Xj3D viewer ( Windows , Mac OS X . and Linux; available at: http://www . web3d . org/x3d/xj3d/ ) . | Most animals , including humans , have sensory and motor capabilities that are biased in the forward direction . The black ghost knifefish , a nocturnal , weakly electric fish from the Amazon , is an interesting exception to this general rule . We demonstrate that these fish have sensing and motor capabilities that are omnidirectional . By combining video analysis of prey capture trajectories with computational modeling of the fish's electrosensory capabilities , we were able to quantify and compare the 3-D volumes for sensation and movement . We found that the volume in which prey are detected is similar in size to the volume needed by the fish to stop . We suggest that this coupling may arise from constraints that the animal faces when using self-generated energy to probe its environment . This is similar to the way in which the angular coverage and range of an automobile's headlights are designed to match certain motion characteristics of the vehicle , such as its typical cruising speed , turning angle , and stopping distance . We suggest that the degree of overlap between sensory and movement volumes can provide insight into the types of control strategies that are best suited for guiding behavior . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"neuroscience",
"computational",
"biology"
] | 2007 | Omnidirectional Sensory and Motor Volumes in Electric Fish |
The nuclear pore complex ( NPC ) is the gate to the nucleus . Recent determination of the configuration of proteins in the yeast NPC at ∼5 nm resolution permits us to study the NPC global dynamics using coarse-grained structural models . We investigate these large-scale motions by using an extended elastic network model ( ENM ) formalism applied to several coarse-grained representations of the NPC . Two types of collective motions ( global modes ) are predicted by the ENMs to be intrinsically favored by the NPC architecture: global bending and extension/contraction from circular to elliptical shapes . These motions are shown to be robust against tested variations in the representation of the NPC , and are largely captured by a simple model of a toroid with axially varying mass density . We demonstrate that spoke multiplicity significantly affects the accessible number of symmetric low-energy modes of motion; the NPC-like toroidal structures composed of 8 spokes have access to highly cooperative symmetric motions that are inaccessible to toroids composed of 7 or 9 spokes . The analysis reveals modes of motion that may facilitate macromolecular transport through the NPC , consistent with previous experimental observations .
Any macromolecule entering or exiting the nucleus must pass through a nuclear pore complex ( NPC ) . NPCs are formed by hundreds of proteins organized into cylindrically symmetric pores that act as gateways to the cell nucleus . The NPC has a mass of ∼50 MDa in yeast and up to 125 MDa in vertebrates , comparable to a small organelle . The yeast NPC is a ring of about 100 nm diameter , creating a central pore of ∼30 nm diameter . It contains approximately 450 proteins , termed nucleoporins , arranged into eight similar “spokes” extending from the NPC's central channel to its outer perimeter ( Figure 1 ) . Each spoke of the yeast NPC exhibits quasi-twofold symmetry about the lumenal plane of the nuclear envelope ( NE ) , giving the NPC quasi-sixteenfold symmetry and reducing the number of unique proteins in the complex to only about 30 . The NPC central channel is coated by several “FG nucleoporins” that characteristically contain multiple structurally disordered phenylalanine ( F ) and glycine ( G ) repeats . The FG nucleoporins collectively serve as a selective barrier between the nucleoplasm and the cytoplasm: Small ( <5–30 kDa ) particles diffuse freely through the NPC channel , but larger particles require the assistance of karyopherin transport factors to pass through the entropic barrier created by the FG nucleoporins . Transport through the NPC is passive , and the rate of material exchange through the NPC can be accounted for by the concentration gradient of the karyopherins and their cargoes [1] . The NPC is somewhat plastic [2] and has been observed to undergo large-scale conformational fluctuations in the form of elongation or dilation under a variety of conditions . Nuclear export of large mRNA molecules seems to be facilitated by dilation of the NPC [3] , [4] . Similarly , some modes of viral infection suggest that the NPC channel widens concurrently with import of viral genetic material [5]: Viral infection often includes the nuclear import of intact capsids , as is the case with papovaviruses [6] , [7] and hepatitis B virus [8] , or import of the intact viral genome , as occurs with adenoviruses [9] and type-1 herpes simplex virus [5] . The NPC also dilates in response to cation concentration [10] and to certain steroids [11] , [12] , possibly as a mechanism to regulate transport rates [11] . Recent studies with cryo-electron tomography [13] , [14] highlight the structural changes of the NPC during transport , indicating that the central channel alters its profile in response to the presence of cargo . Although conventional models of nuclear transport focus primarily on the local interactions of FG nucleoporins with karyopherins [15]–[21] , a number of mechanisms have been proposed to account for the large-scale motions of the NPC [22] , [23] . The limiting factor in determining the nature of the global NPC motions has been the resolution of available data . Cryo-EM and electron tomography images of NPCs reveal global features of the complex , while blurring over its finer details . Alternatively , X-ray crystallography captured the structures of several individual nucleoporins to atomic resolution , but it does not provide the relative positions or orientations of the nucleoporins within the assembled NPC . Recently , an integrative approach has been developed to determine the molecular architectures of macromolecular assemblies from diverse data [24] . Combining spatial restraints that account for the NE excluded volume ( from EM ) , nucleoporin excluded volumes ( from the protein sequences and ultracentrifugation ) , protein positions ( from immuno-EM ) , protein contacts ( from affinity purification ) , and the eight-fold and two-fold symmetries of the NPC ( from EM ) , this approach was used to generate a coarse-grained structural model for the yeast NPC , defining the relative positions of its constituent proteins [25] . Although each individual restraint contains little structural information , the concurrent satisfaction of all restraints from independent experiments drastically reduces the degeneracy of the structural solutions and yields pronounced maxima in the localization probabilities of almost all nucleoporins , thus leading to one predominant NPC architecture [25] . This architecture provides us with enough detail to construct a dynamical model of the NPC at the resolution of a single protein , enabling the exploration of the large-scale dynamics of the NPC . A dynamical model that has enjoyed considerable success in elucidating the machinery of biomolecular assemblies near physiological conditions is the elastic network model ( ENM ) [26] . In the simplest ENM , each of the constituent bodies in the system is represented as a point mass , or node , and a network is constructed by joining neighboring nodes through elastic edges that supply linear restoring forces to displacements from equilibrium . ENMs are straightforwardly paired with normal mode analysis ( NMA ) to provide analytical solutions to motions accessible to biomolecular systems under native state conditions . It has been repeatedly shown that the collectives modes predicted by the ENM correlate well with experimentally observed functional changes in structure triggered by substrate binding or activation [27] . This observation supports the idea that biomolecular systems possess intrinsic dynamics that are uniquely defined by their three-dimensional ( 3D ) structures [27] , in accord with experimental observations [28] . Further , ENMs are readily extensible to large systems through a variety of coarse-graining techniques , and they have been applied at various degrees of detail to systems ranging in scale from small proteins to supramolecular assemblies such as the ribosome [29] , [30] and viral capsids [31] , [32] . The robustness of the global modes predicted by the ENMs and their elegant simplicity and scalability lead us to explore the dynamics of the NPC using ENM-based theory and methods . The known agreement between ENM normal modes and molecular fluctuations can be exploited for structural refinement [33] , [34] . Starting from an initial structure , one can iteratively construct an ENM and modify the structure until the structure matches the observed low-resolution density data . Such techniques are commonplace in generating atomic coordinates to fit cryo-EM data [35]–[37] , and may similarly be used to refine the structure of the NPC . Here , we examine the dynamics of the NPC at several levels of coarse-graining with a newly developed version of a classical ENM . We begin by modeling the NPC on the coarsest scale possible , i . e . , representing the full complex as a flexible toroid . We then increment the level of detail to include non-uniform mass distributions , spokes , and finally the full NPC architecture . We find that the global modes of the NPC are largely captured by the modes of a simple toroid , suggesting that the global dynamics of the NPC does not depend on the details of the NPC architecture . We note that the eight-fold symmetry permits ease of certain cooperative motions that are inaccessible to structures with other rotational symmetries . Two highly robust modes of collective deformation emerge from the analysis , providing insight into the details of NPC motions that are observed with low-resolution tomography [13] , [14] . | The nuclear pore complex ( NPC ) serves as the sole gateway to the cell nucleus , and its proper functioning is therefore crucial for gene expression and many vital signaling pathways . Although it is typically circular , the overall structure of the NPC has been observed to change in response to the presence of cargo . Recently , the molecular architecture of the yeast NPC , including the shapes and relative positions of its constituent proteins , has been resolved . These new structural data provide us with a first opportunity to construct an accurate dynamical model of a macromolecular machine containing hundreds of proteins . By modeling the NPC as a network of masses connected by springs , we investigate its probable large-scale dynamics . We start from a very coarse model and gradually refine it , observing how the structural details influence the calculated dynamics . We find that the NPC dynamics are quite similar to those of a flexible toroid with an uneven mass distribution , and that the 8-fold symmetry that is universally observed in NPCs enables them to undergo certain collective motions that are inaccessible to structures of other symmetries . | [
"Abstract",
"Introduction"
] | [
"biophysics/macromolecular",
"assemblies",
"and",
"machines",
"biophysics/theory",
"and",
"simulation"
] | 2009 | Global Motions of the Nuclear Pore Complex: Insights from Elastic Network Models |
Unlike most eukaryotes , a kinetochore is fully assembled early in the cell cycle in budding yeasts Saccharomyces cerevisiae and Candida albicans . These kinetochores are clustered together throughout the cell cycle . Kinetochore assembly on point centromeres of S . cerevisiae is considered to be a step-wise process that initiates with binding of inner kinetochore proteins on specific centromere DNA sequence motifs . In contrast , kinetochore formation in C . albicans , that carries regional centromeres of 3–5 kb long , has been shown to be a sequence independent but an epigenetically regulated event . In this study , we investigated the process of kinetochore assembly/disassembly in C . albicans . Localization dependence of various kinetochore proteins studied by confocal microscopy and chromatin immunoprecipitation ( ChIP ) assays revealed that assembly of a kinetochore is a highly coordinated and interdependent event . Partial depletion of an essential kinetochore protein affects integrity of the kinetochore cluster . Further protein depletion results in complete collapse of the kinetochore architecture . In addition , GFP-tagged kinetochore proteins confirmed similar time-dependent disintegration upon gradual depletion of an outer kinetochore protein ( Dam1 ) . The loss of integrity of a kinetochore formed on centromeric chromatin was demonstrated by reduced binding of CENP-A and CENP-C at the centromeres . Most strikingly , Western blot analysis revealed that gradual depletion of any of these essential kinetochore proteins results in concomitant reduction in cellular protein levels of CENP-A . We further demonstrated that centromere bound CENP-A is protected from the proteosomal mediated degradation . Based on these results , we propose that a coordinated interdependent circuitry of several evolutionarily conserved essential kinetochore proteins ensures integrity of a kinetochore formed on the foundation of CENP-A containing centromeric chromatin .
The centromeric histone CENP-A acts as the epigenetic mark of a functional centromere from yeast to humans [1] . As a histone H3 variant , CENP-A replaces canonical histone H3 to mark specialized centromeric chromatin by a mechanism that remains largely unknown . While CENP-A deposition occurs in a sequence-dependent manner in point centromeres of certain budding yeasts including S . cerevisiae [2] , [3] , its recruitment to regional centromeres of most other eukaryotes is not strictly sequence dependent [1] , [3]–[5] . Several lines of evidence suggest that the composition of nucleosomes that form centromeric chromatin may vary from species to species [6]–[10] . Even the hierarchy of events that assembles and stabilizes a complex macromolecular kinetochore ( KT ) structure on unusual centromeric chromatin , whether universal or species-specific , remains unclear . CENP-A is believed to be the initiator of the process of KT formation [1] , [11] . Localization of most KT proteins is regulated by CENP-A [12]–[15] . However , a few proteins such as Ndc10 and Scm3 in S . cerevisiae [6] , [16] , Mis12/Mtw1 in C . albicans [17] , Mis6 , Mis16-Mis18 complex and Ams2 in Schizosaccharomyces pombe [18] , [19] , and CENP-H-I complex in humans [20] have been shown to regulate CENP-A localization . Although the structure of metazoan KTs can be visualized under microscope , the exact nature of the KT architecture is difficult to ascertain in yeasts due to small size of these cells [21] . However , based on presence of functional homologs of several KT proteins , and their genetic as well as biochemical interaction in various eukaryotes , it is presumed that the three-layered KT structure is evolutionarily conserved from yeasts to humans . A KT helps in bridging the mitotic spindle to centromere ( CEN ) DNA to ensure faithful chromosome segregation during mitosis and meiosis . KT proteins exist as sub-complexes that assemble on CEN DNA [22]–[24] . In humans , inner ( CENP-A , -B , -C , -H and –I ) and middle ( the Spc105 complex and the Mis12 complex ) KT proteins are associated constitutively with CEN DNA [25] but outer KT proteins ( such as the Ndc80 complex and the Ska1 complex ) which help in KT-microtubule ( MT ) interaction are generally localized to KTs only during mitosis [26] , [27] . In contrast , KTs are fully assembled in S . cerevisiae early in the cell cycle . The fungal specific Dam1 complex , an outer KT protein complex in budding yeast S . cerevisiae , remains associated with KTs throughout the cell cycle [28]–[30] . One of the less understood features of budding yeast KTs is that they are clustered together throughout the cell cycle [31] . A recent study , using chromosome conformation capture-on-chip ( 4C ) , clearly demonstrates that all chromosomes cluster via centromeres at one pole of the nucleus in S . cerevisiae suggesting interchromosomal cross-talks through inter-KT interaction [32] . KT clustering has been shown to be important for centromere function in S . cerevisiae as KTs are found to be declustered in ndc10 , ame2 and nuf2 KT mutants [31] , [33] . Interestingly , KTs are clustered only during interphase but not in mitosis in fission yeast S . pombe [34] . In metazoans KTs are never clustered [35] . With the notable exception of holocentric chromosomes of nematodes and aphids where KTs are formed across the entire length of a chromosome [36] , only one KT is formed on each chromosome in all other organisms studied till date . A functional KT can even assemble on non-centromeric DNA to form a neocentromere in certain organisms when a native centromere is deleted or inactivated [37]–[40] . Thus there must be an underlying active mechanism to prevent formation of centromeric chromatin on neocentromeric loci when the native centromere is functional . CENP-A at non-centromeric locations is targeted for proteasomal degradation in Drosophila melanogaster and S . cerevisiae [41]–[43] . Thus ectopic CENP-A is destabilized to prevent multiple kinetochore formation although the exact cellular signal that distinguishes CENP-A molecules present at the native centromere from those ectopically localized could not be determined . Candida albicans , a pathogenic budding yeast , that causes candidiasis in humans , carries 3–5 kb long unique centromeric chromatin on each of its eight chromosomes . There are 4 CENP-A molecules but only one MT binds to a KT in C . albicans [44] . Centromeric regions have been shown to have unusual chromatin structure [45] and histone H3 molecules are replaced by CENP-A at CENs ( K . Sanyal , unpublished ) in this organism . Moreover , CENP-A deposition on CENs has been shown to be epigenetically regulated [45]–[47] . We have previously cloned and characterized several evolutionarily conserved KT proteins including CENP-A/Cse4 [48] , CENP-C/Mif2 [46] , Mis12/Mtw1 [17] and the Dam1 complex subunits [49] in C . albicans and shown that each of these proteins is essential in a KT-MT-mediated process of chromosome segregation . In the present work , we studied localization interdependence of KT proteins that govern KT integrity including stability of CENP-A . Our results reveal that the KT architecture is stabilized in a coordinated interdependent manner by individual components of the KT in C . albicans . Most strikingly , we provide evidence that stability of CENP-A molecules is determined by integrity of the KTs . This is the first demonstration , to our knowledge , of how an interdependent circuitry of several KT proteins helps stabilizing CENP-A at a functional KT .
In order to understand the process of KT assembly in C . albicans , we studied localization dependence of various essential KT proteins that are evolutionarily conserved . To achieve this , we utilised conditional mutant strains carrying KT proteins under control of the MET3 or PCK1 promoter . The MET3 promoter is repressed in presence of cysteine ( Cys ) and methionine ( Met ) [50] while the PCK1 promoter is repressed when glucose ( Glu ) is used as the carbon source [51] . We depleted each of Dam1 , Ask1 , Spc19 or Dad2 - subunits of an essential outer KT protein complex , the Dam1 complex , in J102 ( MET3prDAM1/dam1 ) , J104 ( MET3prASK/ask1 ) , J106 ( MET3prSPC19/spc19 ) or J108 ( PCK1prDAD2/dad2 ) [49] respectively to study localization dependence of CENP-A . Immunostaining with anti-Cse4 ( CENP-A ) antibodies in depleted levels of each of these subunits of the Dam1 complex revealed that CENP-A localization at the KTs was dramatically reduced as compared to conditions when these proteins were present at wild-type levels ( Figure 1A , Figure S1 ) . In order to examine whether CENP-A delocalization is a specific phenomenon associated with depletion of the Dam1 complex , present only in fungal kingdom , we sought to study CENP-A localization upon depletion of the homolog of Nuf2 , another evolutionarily conserved outer KT protein , in C . albicans . Depletion of Nuf2 in YJB12326 ( MET3prNUF2/nuf2 ) showed significantly reduced CENP-A localization as well ( Figure 1A ) . Next , we examined localization of CENP-A in absence of CENP-C/Mif2 in C . albicans . CENP-C/Mif2 proteins are evolutionarily conserved inner KT proteins . CENP-A/Cse4 staining in CAMB2 ( PCK1prMIF2/mif2 ) grown in repressive media ( Glu ) for 8 h revealed a significant loss of CENP-A/Cse4 from KTs ( Figure 1A ) . Thus various proteins that are evolutionarily conserved and present at inner , middle and outer KT in many organisms influence CENP-A localization in C . albicans . Next , to study localization patterns of CENP-C/Mif2 in absence of outer KT proteins , we performed immunostaining using anti-Myc antibodies in J123 ( MET3prDAM1/dam1 MIF2/PCK1pr12XMYCMIF2 ) and J124 ( MET3prASK1/ask1 MIF2/2PCK1pr12XMYCMIF2 ) expressing Myc-tagged CENP-C/Mif2 from the PCK1 promoter . Similar to CENP-A ( discussed above ) , CENP-C/MycMif2 localization was dramatically reduced when levels of Dam1 or Ask1 were depleted due to growth of J123 or J124 for 8 h under non-permissive conditions ( +Cys +Met +Suc ) of the MET3 promoter ( Figure 1B ) . Recently , we demonstrated that in C . albicans KT occupancy of these two proteins is dependent on the cellular levels Mis12/Mtw1 , an evoutinarily conserved middle KT protein [17] . Thus , assembly of the inner KT is dependent on proteins from middle and outer KT in C . albicans . These results prompted us to further investigate the role of the Dam1 complex on the occupancy of a middle KT protein . KT localized Mtw1-GFP signals that were visible in wild-type conditions were absent when Ask1 or Dam1 was repressed for 8 h in J122 ( MET3prDAM1/dam1 MTW1GFP/MTW1 ) or J120 ( MET3prASK1/ask1 MTW1GFP/MTW1 ) ( Figure 1C ) . Together we conclude that integrity of the middle KT is also determined by outer KT proteins . Having established localization dependence of various domains of a KT on each other , we further examined how localization of one protein depends on another at the outer KT . As compared to wild-type , Nuf2-GFP localization at the KT was reduced dramatically when Dam1 was depleted in YJB12289 ( MET3prDAM1/dam1 NUF2GFP/NUF2 ) ( Figure 1D ) . These results suggest that integrity of different domains of a KT is interdependent and assembly of various components of a KT is highly coordinated ( Figure 1E ) . Unlike most organisms including fission yeast and humans , KTs are attached to spindle MTs throughout the cell cycle in S . cerevisiae except for a brief period of 2–3 minutes during centromere replication [52] , [53] . A similar KT-MT interaction at all stages of the cell cycle is evidenced in C . albicans as well [17] . Various proteins from C . albicans discussed above have been shown to be essential in KT-MT mediated process of chromosome segregation as severe spindle defects were observed upon depletion of each of these proteins [17] , [46] , [48] , [49] . In this study , we examined whether or not reduced KT localization of various KT proteins was due to impairment of the mitotic spindle caused by depletion of each of these proteins . To test this possibility , we disrupted the mitotic spindle in 10118 ( CSE4:GFP:CSE4/cse4 ) expressing GFP-tagged Cse4 by treating cells with a spindle depolymerizing drug nocodazole ( NOC ) . Tubulin staining of these NOC treated cells exhibited disruption of the mitotic spindle structure as expected ( Figure 2A ) . However , no significant change in the intensity of Cse4-GFP signals was observed between NOC treated ( mean intensity value = 225±28 a . u . ) and untreated ( mean intensity value = 227±32 a . u . ) cells of 10118 ( Figure 2B , Figure S2 ) . A similar experiment to compare Mtw1-GFP levels in NOC treated and untreated cells of YJB10695 ( MTW1GFP/MTW1 ) exhibited no significant difference ( Figure 2C ) as well . Localization of Dad2 , a subunit of the Dam1 complex , is unaltered in presence or absence of NOC [49] . Together , these results showed that localization of various components of a KT is independent of integrity of the mitotic spindle . CENP-A/Cse4 ChIP analysis further indicated that spindle integrity does not have significant effect on the stability of centromeric chromatin ( Figure 2D ) . In order to understand how absence of a KT protein leads to collapse of an entire KT , we examined the KT structure at reduced ( partially depleted ) levels of various KT proteins . We observed both CENP-A ( Cse4 ) and CENP-C ( MycMif2 ) signals in wild-type or partially depleted levels of Dam1 or Ask1 in J123 ( MET3prDAM1/dam1 MIF2/PCK1pr12XMYCMIF2 ) or J124 ( MET3prASK1 ) /ask1 MIF2/2PCK1pr12XMYCMIF2 ) . Interestingly , after 4–5 h of Dam1 depletion we observed multiple weak signals of CENP-A ( Cse4 ) and CENP-C ( MycMif2 ) associated with a single nucleus as opposed to a single bright dot-like clustered KTs observed in each wild-type cell ( Figure 3A ) . Ask1-depleted cells showed similar declustered KTs ( data not shown ) . We also observed multiple CENP-A ( Cse4 ) signals in partially depleted Dad2 cells of J108 ( PCK1prDAD2/dad2 ) after 4–5 h of growth under non-permissive conditions ( Figure S3A ) . Subsequently , we examined Mtw1-GFP signals by partially depleting Ask1 in J120 ( MET3prASK1/ask1 MTW1GFP/MTW1 ) . We observed multiple Mtw1-GFP signals per nucleus in these cells as well ( Figure S3B ) . Nuf2-GFP showed localization patterns in Dam1 depleted conditions ( data not shown ) . To test whether KT disintegration occurs due to depletion of middle ( Mis12/Mtw1 ) or inner ( CENP-A/Cse4 ) KT proteins as well , we first depleted Mis12/Mtw1 in CAKS12 ( PCK1prMTW1/mtw1 ) , and examined the integrity of the clustered KTs using anti-Cse4 antibodies . KT disinegration was clearly evident with multiple weak CENP-A signals co-localized with a single nucleus ( Figure 3B ) . Next we monitored the process of Mis12/Mtw1 delocalization upon depletion of CENP-A . YJB11483 ( PCK1prCSE4/cse4 MTW1GFP/MTW1 ) expressing CENP-A/Cse4 under the PCK1 promoter was grown either in permissive ( +Suc ) or non-permissive ( +Glu for 5 h ) conditions to examine Mtw1-GFP signals ( Figure 3C ) . We observed multiple Mtw1-GFP signals per cell confirming disintegration of the KT cluster in these cells as well ( as opposed to wild-type cells where properly integrated KTs remained clustered ) . Using the LSM examiner analysis tool ( Carl Zeiss , Germany ) , we determined the mean intensity of GFP signals in wild-type clustered KTs and Dam1 depleted declustered KTs ( Figure 4 ) . The normalized ( against background ) average values of mean GFP intensity/KT in wild-type and Dam1 depleted cells were 9±1 . 8 and 1 . 5±0 . 29 respectively . Thus the Mtw1 levels at the disingrated KTs are significantly reduced due to Dam1 depletion . Finally , we examined the dynamics of KT disassembly by monitoring the time-dependent localization patterns of several proteins present at various domains of a KT while Dam1 is being gradually depleted . Cse4-GFP , Mtw1-GFP and Nuf2-GFP signals were monitored in J127 ( MET3prDAM1/dam1 CSE4:GFP:CSE/cse4 ) , J122 ( MET3prDAM1/dam1 MTW1/MTW1GFP ) and YJB12289 ( MET3prDAM1/dam1 , NUF2GFP/NUF2 ) respectively at various time intervals upon Dam1 depletion ( Figure 5 ) . In each case , disintegration of GFP signals from one bright cluster to multiple weakly fluorescent dot-like signals in each cell was observed between 4–5 h of growth in Dam1 repressing media . GFP signals were undetectable after 8 h of Dam1 depletion indicating complete collapse of KT integrity . Similar disintegration dynamics of Mtw1-GFP signals were found between 4–5 h of depletion of other proteins , CENP-A or Ask1 , as well ( Figure S4 ) . These results indicated a strong correlation between a concerted loss of various domains of a KT and a concomitant reduction in the levels of an essential KT protein . All KTs are clustered and such clustered KTs are always localized towards the nuclear periphery ( Figure 6 , a-a″ ) suggesting KTs occupy a fixed nuclear territory in wild-type C . albicans cells . Interestingly , reconstruction of three dimensional ( 3D ) images revealed declustered KT signals resulting due to Dad2 depletion in J108 ( PCK1prDAD2/dad2 ) retained nuclear peripheral localization . Since peripheral nuclear localization of declustered KTs is maintained ( Figure 6 , b-b″ , c-c″ ) , we speculate that KTs , whether clustered or not , remain largely attached to nuclear periphery through some components that are not affected due to depletion of an individual KT protein . To examine the status of centromeric chromatin when KT integrity is lost , ChIP assays with anti-Cse4 antibodies were performed in BWP17 ( wild-type ) , J102 ( MET3prDAM1/dam1 ) and J108 ( PCK1prDAD2/dad2 ) strains grown in non-permissive media for 8 h . These experiments revealed a drastic decrease in CENP-A binding to CENs in Dam1 or Dad2 depleted cells as compared to wild-type cells confirming that integrity of CENP-A-bound centromeric chromatin is highly affected in these mutants ( Figure 7A ) . Since the localization of CENP-C/Mif2 was also affected when Dam1 was depleted we tested centromere occupancy of CENP-C/Mif2 in J125 ( MIF2/PCK1pr12XMYCMIF2 ) or depleted levels of Dam1 in J123 ( MET3prDAM1 ) /dam1 MIF2/PCK1pr12XMYCMIF2 ) by ChIP assays with anti-MycMif2 antibodies ( Figure 7B ) . The ChIP-PCR analysis confirmed that CENP-C/Mif2 occupancy at centromeres was also dramatically reduced in Dam1 depleted cells as compared to wild-type cells . We have recently shown that , Mtw1/Mis12 is required for inner kinetochore assembly including localization of CENP-A and CENP-C [17] . Together these results confirmed that Dam1 is required for integrity of centromeric chromatin in C . albicans . Since centromeric chromatin is disintegrated when various KT proteins are depleted , we examined protein levels of CENP-A in these conditions to find out the fate of CENP-A molecules that are no longer associated with centromeres . We prepared lysates from J108 ( PCK1prDAD2/dad2 ) grown overnight in Suc ( expressed condition ) or various time intervals after transferring in Glu ( repressed condition ) , and performed western blot analysis to measure the levels of both Dad2 and CENP-A . A decrease in Dad2 protein levels in cells grown at increasing time in Glu confirmed time-dependent repression of Dad2 expression by the PCK1 promoter ( Figure 8A , top panel ) . Strikingly , a concomitant reduction in CENP-A levels , as determined by western blot analysis using anti-Cse4 antibodies , with decreasing levels of Dad2 was also observed ( Figure 8A , bottom panel ) . To examine the fate of CENP-A in reduced levels of other subunits of the Dam1 complex , total cell lysates were prepared from strains where Dam1 ( J102 ) , Ask1 ( J104 ) , or Spc19 ( J106 ) was either expressed or repressed for 8 h . As observed in Dad2-depleted cells , CENP-A levels in the total cell lysate were found to be highly reduced in absence of each of these proteins ( Figure S5 ) . Next we examined CENP-A stability in depleted conditions of CENP-C/Mif2 , Mis12/Mtw1 , and Nuf2 - evolutionarily conserved inner , middle and outer KT proteins respectively . Total cell lysate was prepared from each sample collected at different time points during Mif2/CENP-C , Mis12/Mtw1 or Nuf2 reprefssion in CAMB2 ( PCK1prMIF/mif2 ) , CAKS12 ( PCK1prCSE4/cse4 ) or YJB12326 ( MET3prNUF2/nuf2 ) . Western blot analysis of these cell lysates with anti-Cse4 antibodies confirmed a decrease in CENP-A levels when either CENP-C ( Figure S5 ) , Mis12 or Nuf2 ( Figure 8B ) was depleted . Thus , CENP-A protein stability is dependent on wild-type levels of several KT proteins . To investigate whether increased levels of CENP-A could rescue KT integrity in depleted levels of a KT protein we sought to express a mutant stable form of CENP-A in C . albicans . In budding yeast S . cerevisiae Cse4/CENP-A at non-centromeric regions is degraded by the proteasomal mediated degradation pathway which is partially prevented by replacement of all lysine residues with arginine residues in ScCse4 ORF . Using site-directed mutagenesis , all seven lysine residues in Cse4 ORF in C . albicans strain ( CAKS2b ) were replaced by arginine residues to construct the strain J129 ( CSE47R-Prot A/cse4 ) where Protein A ( Prot A ) tagged mutated Cse47R is the only source of CENP-A ( Figure 9A ) . All the changes were confirmed by DNA sequencing ( Figure S6 ) . Western blot analysis ( Figure 9B ) and immunolocalization ( Figure 9C ) using anti-Prot A antibodies revealed that Cse47R -Prot A is functionally expressed in C . albicans . To examine the effect of Dam1 depletion on the stability of Cse47R , we expressed Cse4-Prot A and Cse47R-Prot A in the Dam1 conditional mutant . Western blot analysis in J130 ( MET3prDAM1 ) /dam1 CSE4-Prot A/CSE4 ) or J131 ( MET3prDAM1/dam1 CSE47R-Prot A/CSE4 ) revealed that while wild-type Cse4-Prot A was degraded , Cse47R -Prot A remained stable upon Dam1 depletion ( Figure 10A ) . This suggests that the proteasomal mediated pathway is involved in degradation of unbound CENP-A in C . albicans . To test whether increased levels of CENP-A ( Cse47R ) could rescue KT disintegration caused by depletion of Dam1 , we analysed Cse47R localization at the KTs in Dam1 mutant . Immunolocalization using anti-Cse4 antibodies in J130 ( MET3prDAM1 ) /dam1 CSE4-Prot A/CSE4 ) or J131 ( MET3prDAM1/dam1 CSE47R-Prot A/CSE4 ) cells revealed that similar to wild-type Cse4 , Cse47R failed to localize at the KTs in absence of Dam1 ( Figure 10A ) . Prot A ChIP analysis from J131 ( MET3prDAM1/dam1 CSE47R-ProtA/CSE4 ) cells showed a significant drop in Cse47R binding to centromeres of all chromosomes under Dam1 depleted conditions . Thus Dam1 depletion leads to disintegration of centromeric chromatin containing wild-type CENP-A ( Cse4 ) and stable form of CENP-A ( Cse47R ) in the same manner ( Figure 10B ) . We next examined whether newly synthesized CENP-A molecules can be recruited to the KT once the process of KT disassembly is initiated under Dam1 depleted conditions . We utilized YJB11990 ( PCK1prCSE4/CSE4 MET3prDAM1/dam1 ) in which CSE4 and DAM1 are placed under control of the PCK1 and MET3 promoters respectively . We studied the fate of newly synthesized CENP-A molecules by inducing expression of CENP-A/Cse4 ( +Suc ) while Dam1 is kept in repressed state ( +Cys +Met ) . To achieve this , YJB11990 was grown in Dam1 repressible conditions till the point where ( a ) KTs starts to decluster ( 4 h ) or ( b ) KTs are disintegrated ( 8 h ) ( Figure 11A ) . These cells were then transferred to media that represses the MET3 promoter ( +Cys +Met to keep Dam1 depleted ) but induces the PCK1 promoter ( +Suc to overexpress Cse4 ) and grown for an additional 4 h . Expression of new CENP-A/Cse4 molecules in such conditions was confirmed by western blot analysis ( Figure 11A ) . Indirect immunolocalization using anti-Cse4 antibodies revealed that newly synthesized CENP-A/Cse4 molecules expressed from the induced PCK1 promoter ( Figure 11A ) could not be recruited at KTs under Dam1 repressible conditions ( Figure 11B ) .
CENP-A is required for localization of many KT proteins [12] , [14] , [15] . In S . cerevisiae , inner and middle KT proteins ( Ndc10 , Mif2 , Mtw1 , and Okp1 ) showed 50% decrease in occupancy at the active conditional centromere ( cCEN ) in absence of CENP-A/Cse4 , whereas the middle ( Ctf19 ) and many outer ( Ndc80 , Dam1 , Ask1 , and Stu2 ) KT proteins completely failed to localize [43] . This suggests that CENP-A/Cse4 is the initiator of KT assembly . However , several kinetochore proteins do not require CENP-A for localization , suggesting that CENP-A independent species-specific KT assembly pathways also exist in certain organisms [15] , [18] , [56] . CENP-A localization has been shown to influence Mis12 localization in S . cerevisiae , D . melanogaster and humans but not in S . pombe [55] , [57]–[59] . On the other hand , Mis12 does not influence localization of CENP-A in most organisms except in C . albicans where CENP-A and Mis12 localization is interdependent [15] , [55] , [58] , [59] . Depletion of CENP-A affects CENP-C localization in S . cerevisiae , C . elegans and humans but CENP-C has no effect on CENP-A localization in these organisms [13] , [14] . In this study , we examined how CENP-A localization is influenced by other KT proteins in C . albicans . Intriguingly , centromere localization of CENP-A was dramatically reduced due to depletion of inner ( Mif2/CENP-C ) or outer ( Dam1 complex and Nuf2 ) KT proteins . Recently , we showed that a middle KT protein Mis12/Mtw1 homolog C . albicans influences assembly of two inner KT proteins CENP-A and CENP-C [17] . Thus localization dependence of CENP-A on various KT proteins varies from species to species . Since it is unusual and striking that outer KT proteins influence localization of CENP-A , we further investigated localization dependence of various other proteins to determine the sequence of events that lead to KT formation on unique short regional centromeres of C . albicans . An unprecedented observation of collapse of the KT architecture in absence of an essential protein from a KT in C . albicans raises an important question . How do KTs assemble in C . albicans ? KT disassembly due to depletion of various proteins suggests that KT assembly is probably not a step-wise process in C . albicans . We propose that KT proteins of various sub-complexes assemble in a unique interdependent concerted manner to form and stabilize the macromolecular KT architecture in C . albicans Our results thus indicate that the KT in C . albicans may not even have a layered structure , unlike the one observed in humans . Since all proteins we analyzed in this study have been previously shown to be essential for the proper dynamics of a mitotic spindle in C . albicans [17] , [48] , [49] we investigated whether or not an intact mitotic spindle is required for maintaining integrity of KTs . Localization of CENP-A or Mis12 was found to be unaffected in presence of a MT-depolymerizing drug nocodazole . In NOC treated cells tubulin staining showed two weak dot-like signals probably representing SPBs after duplication . Cse4 and Mtw1 GFP signals in NOC treated cells are also seen as two dots situated close to each other . Thus we conclude that stabilization of the KT structure is independent of structural integrity of the mitotic spindle . Like S . cerevisiae , KT-MT interaction is established early during S phase in C . albicans . Live cell imaging by time-lapse microscopy in our previous study revealed that all centromeres are clustered together throughout the cell cycle in C . albicans [17] . Moreover , similar localization patterns of these clustered centromeres at the nuclear periphery were observed both in S . cerevisiae and C . albicans . In absence of a metaphase plate in budding yeasts [60] , clustered centromeres may provide a platform for MT attachment and synchronous separation of sister chromatids during anaphase . Conditional KT mutants provided us an opportunity to follow the process of KT disassembly in a time dependent manner during gradual depletion of various KT proteins in C . albicans . Our results clearly indicate that disintegration of the KT cluster is a common intermediate step before KT collapse and each component is required to keep all the KTs clustered in C . albicans . Chromosome confirmation capture ( 3C ) assay to study interaction among centromeres of different chromosomes in S . cerevisiae revealed that crosslinking frequencies between different centromeres is significantly higher than other chromosomal sites except telomeric regions [61] . A recent study , where chromosome confirmation capture on chip ( 4C ) and massively parallel sequencing were used to globally capture inter and intrachromosomal interactions in S . cerevisiae , demonstrates centromeres as the chromosome landmarks that mediate interchromosomal interaction [32] . The clustering of centromeres that marked the primary point of engagement between different chromosomes was the most striking feature of the inter-chromosomal contacts . In the light of these observations in S . cerevisiae and our results in this study , we speculate that interchromosomal interaction at the centromeres may facilitate stabilization of the kinetochore ensemble as depletion of various KT proteins leads to disintegration of clustered centromeres in both species . ChIP analysis and immunolocalization studies confirmed that CENP-A occupancy is severely impaired when an essential KT protein is depleted from C . albicans cells . We predicted that disintegration of centromeric chromatin in absence of KT proteins may expose free CENP-A molecules for cellular degradation . Indeed , western blot analysis confirmed that cellular CENP-A protein levels were drastically reduced in KT mutants tested . Since integrity of centromeric chromatin is also dependent upon individual KT proteins which in turn help in maintaining overall KT integrity , it is tempting to speculate that establishment of centromeric chromatin and assembly of KTs start together in a coordinated way in C . albicans . The ubiquitin mediated degradation pathway is one of the major pathways that degrade CENP-A/Cse4 in S . cerevisiae and CENP-A/CID in Drosophila . In Dam1 depleted strain where wild-type Cse4 is unstable , the mutant Cse47R showed higher stability suggesting a similar CENP-A degradation pathway is active in C . albicans . Thus , although the process of CENP-A recruitment on point and regional centromeres may differ but the mechanism that regulates CENP-A stability to prevent ectopic KT formation ( especially in C . albicans where efficiency of neocentromere formation is remarkably high ) seem to be conserved among organisms carrying point and regional centromeres . Wild-type CENP-A is unstable when the KT ensemble is disintegrated and centromeric chromatin is disrupted in absence of an essential KT protein . It is possible that reduction in CENP-A levels is the cause not the consequence of KT disassembly under such conditions . We wondered whether expression of a stable form of CENP-A ( Cse47R ) can prevent disintegration of the KT structure in absence of Dam1 . We examined stable mutant CENP-A , Cse47R localization at the KT in Dam1 depleted cells and found that increasing the stability of CENP-A ( confirmed by western blot analysis ) does not prevent CENP-A delocalization at the KT in absence of Dam1 . The Cse47R failed to maintain integrity of centromeric chromatin as well . Next we sought to provide new CENP-A molecules using an inducible promoter ( PCK1 ) to examine if newly synthesized CENP-A can be recruited to the KTs in absence of Dam1 . However , these newly synthesized CENP-A molecules ( detected by western blot analysis ) failed to rescue KT integrity further confirming that an individual KT protein is absolutely essential for protecting the centromeric bound CENP-A by maintaining the integrity of the KT ensemble that is laid on the foundation of CENP-A associated centromeric chromatin . Figure 12 shows a possible pathway of KT destabilization due to depletion of an essential KT protein in C . albicans .
Strains and primers used in this study are listed in Table S1 and Table S2 respectively . Conditional mutant strains of Dam1 ( J102 , J121 , J122 , J127 , J130 , J131 , YJB11990 and YJB12289 ) , Ask1 ( J104 , J120 , and J124 ) , Spc19 ( J106 ) and Nuf2 ( YJB12326 ) that carry DAM1 , ASK1 SPC19 and NUF2 respectively under control of the MET3 promoter were grown in YPDU ( 1% yeast extract , 2% peptone , 3% glucose and 0 . 01% uridine ) as permissive media and YPDU+5 mM cysteine ( +Cys ) +5 mM methionine ( +Met ) as non-permissive media . Conditional mutant strains of Mif2 ( CAMB2 , J123 , J124 and J125 ) , Dad2 ( J108 ) , Cse4 ( CAKS3b and YJB11483 ) and Mtw1 ( CAKS12 ) that carry MIF2 , DAD2 , CSE4 and MTW1 respectively under control of the PCK1 promoter were grown in YPSU ( 1% Yeast Extract , 2% Peptone , 2% Succinate and 0 . 01% Uridine ) as permissive media and YPDU as non-permissive media . All C . albicans strains were grown at 30°C . C . albicans conditional mutants with GFP tagged KT proteins grown overnight in inducing media were transferred to repressible media at initial OD600 - 0 . 150 . Cells were harvested at various time intervals after growth in repressible media . Harvested cells were resuspeneded in 50% glycerol and subsequently imaged using a confocal microscope ( Zeiss LSM 510 META ) . Images were further processed by Adobe Photoshop . For nocodazole treatment , cells were grown overnight in YPDU , reinoculated in YPDU with an initial OD600 = 0 . 2 . Nocodazole ( Sigma , Cat # M1404 ) was added at a concentration of 20 µg/ml when OD600 = 0 . 4 ( 1 generation ) was achieved . Cells were grown for an additional 4 h before harvesting for immunolocalization and ChIP assays . ChIP assays were performed using a protocol described previously [49] . An exponentially growing culture of C . albicans strain was fixed with 1% formaldehyde for 15 min . The reaction was quenched for 5 min at room temperature using glycine to a final concentration of 125 mM . Cells were washed and suspended in resuspension buffer ( 0 . 2 mM Tris-HCl pH , 9 . 4 , 10 mM DTT ) . Resuspended cells were incubated at 30°C for 15 min on a shaker at 180 rpm . Cells were washed and resuspended in spheroplasting buffer ( 1 . 2 M Sorbitol , 20 mM Na-HEPES , pH 7 . 5 ) . Spheroplasting ( 95% ) was performed using lyticase ( Sigma , Cat # L2524 ) at 30°C at low speed . Spheroplasting was stopped by adding ice-cold postspheroplasting buffer ( 1 . 2 M Sorbitol , 1 mM MgCl2 , 20 mM Na-PIPES , pH 6 . 8 ) . Spheroplasts were subsequently washed with ice-cold 1× PBS , Buffer I ( 0 . 25% TritonX-100 , 10 mM EDTA , 0 . 5 mM EGTA , 10 mM Na-HEPES , pH 6 . 5 ) , Buffer II ( 200 mM NaCl , 1 mM EDTA , 0 . 5 mM EGTA , 10 mM Na-HEPES ) and finally resuspended in extraction buffer ( 140 mM NaCl , 1 mM EDTA , 50 mM K-HEPES , 0 . 1% sodium deoxycholate , 1% Triton X-100 , pH 7 . 5 ) with protease inhibitor cocktail ( Sigma ) at a concentration of 100 µl/100 ml starting culture . Next , sonication was performed to get sheared chromatin fragments of an average size of 300–700 bp by SONICS Vibra cell sonicator . The soluble fraction of sheared chromatin was obtained by centrifuging the sonicated solution at 13000 rpm for 15 min at 4°C . Indirect immunofluoroscence was performed using protocol described previously [49] . Asynchronous and exponentially growing culture was fixed using 1 ml 37% formaldehyde per 10 ml culture for 1 h . Fixed cells were washed and resuspended in 0 . 1 M Phosphate buffer ( pH 6 . 4 ) . Next , 70–80% spheroplasting was achieved using lyticase ( Sigma ) and β-mercaptoethanol . Spheroplasts were pelleted down gently at low speed and resuspended in PBS . Teflon coated slide was incubated with polylysine ( 1 mg/ml ) for 5 min , washed with water and dried . Next 15 µl of fixed cells were placed onto each well and incubated for 5 min . Cells attached to the slides were fixed in ice cold methanol ( −20°C ) for 6 min and ice-cold acetone ( −20°C ) for 30 seconds . Blocking was performed with 2% skim milk in PBS for 30 min . Subsequently cells were incubated with primary antibodies for 1 h and washed with PBS four times . Subsequently , secondary antibodies were added onto each well and incubated 1 h in a dark humid chamber . Finally the slide was washed four times with phosphate buffered saline ( PBS ) . DAPI solution ( 50 ng/ml in 70% glycerol ) was added on to each well and coverslip and slide were sealed together . Co-immunolocalization experiments were performed using the same protocol . Images were captured using Carl Zeiss confocal laser scanning microscope ( LSM 510 META ) using LSM 5 Image Examiner . Three dimentional ( 3D ) images were generated using LSM 3D rendering software ( Figure 6 and Figure 12 ) ( Carl Zeiss , Germany ) . Images were rotated in 3D and snapshots were taken from three different rotational angles ( a-a″ , b-b′ , c-c″ ) in Figure 6 . Images were susequently processed in Adobe Photoshop . Wild-type or conditional mutant strains were grown under inducing and repressed conditions for 8 h . Protein extracts were made by disrupting the cells in RIPA buffer ( 300 mM NaCl , 50 mM Tris-HCl pH 8 . 0 , 5 mM EDTA pH 8 . 0 , 0 . 5% Triton-X ) using glass beads ( Sigma cat # G8772 ) . The lysate were subjected to electrophoresis using 12% SDS PAGE and transferred to nitrocellulose membrane for 1 h at 20 V by semi-dry method . Proper transfer was checked by Ponceau S staining . Membranes were blocked with 5% skim milk for 1 h followed by incubation with primary antibodies in 5% skim milk overnight at 4°C . Membranes were washed five times with PBS+0 . 05% Tween and incubated with secondary antibodies in 5% skim milk for 2 h . Membranes were washed five times with 1× PBS+0 . 05% Tween and developed by VersaDoc ( Bio-Rad ) or exposed to X-ray films . Quantification of the western blots was performed using the Quantity one software ( Bio-Rad ) . Primary antibodies used for immunolocalization studies were as follows- affinity purified rabbit anti-Dad2 antibodies-1∶50 dilution , affinity purified rabbit anti-CENP-A antibodies - 1∶500 dilution [48] , mouse anti-Myc -1∶50 dilution ( Calbiochem , Cat # OP10L ) , rabbit anti-Prot A- 1∶1500 ( Sigma Cat # P2921 ) , rat anti-tubulin ( Invitrogen , Cat # YOL1/34 ) - 1∶100 dilution . The fluorescent secondary antibodies for immunolocalization were obtained from Invitrogen and used at dilution 1∶500 for Alexa Fluor goat anti-rabbit IgG 568 ( Cat #A11011 ) , 1∶100 for Alexa Fluor goat anti-rat IgG 488 ( Cat # A11006 ) and 1∶500 for Alexa Fluor anti -mouse 488 ( Cat # A11001 ) . Primary antibodies used for western blot analysis were rabbit anti-CENP-A ( 1∶500 ) , mouse anti-PSTAIRE ( 1∶2000 , Sigma Cat # P7962 ) , rabbit anti-Prot A ( 1∶ 5000 , Sigma Cat # P2921 ) and rabbit anti-Dad2 ( 1∶500 , unpurified sera ) antibodies . Secondary antibodies used were anti-rabbit HRP conjugated ( 1∶2000 , Bangalore Genei Cat # 105499 ) , anti-mouse HRP conjugated ( 1∶2000 , Bangalore Genei , Cat # HP06 ) . | The kinetochore , a macromolecular protein complex that assembles on centromere DNA , interacts with spindle microtubules to mediate faithful chromosome segregation . The sequence of centromere DNA , which varies from 125 bp long point centromere in Saccharomyces cerevisiae to a few Mb long regional centromeres in humans , does not show any conservation across species . Kinetochore proteins , however , share a higher degree of conservation in amino acid sequence . Intriguingly , kinetochore assembly has been shown to be species-specific , although specialized centromeric chromatin is always formed by the centromeric histone CENP-A . We investigated this process of kinetochore assembly on epigenetically determined regional centromeres in the pathogenic budding yeast Candida albicans . Here we established that a coordinated interdependent assembly of several essential evolutionarily conserved kinetochore proteins ensures integrity of a functional kinetochore . Depletion of an essential kinetochore protein leads to total collapse of the kinetochore architecture . We observe that kinetochore disintegration precedes kinetochore collapse . Finally , we prove that kinetochore integrity keeps centromeric chromatin intact and protects CENP-A molecules from proteasomal mediated degradation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"evolutionary",
"biology",
"centromeres",
"microbiology",
"mitosis",
"model",
"organisms",
"cell",
"division",
"dna",
"chromatin",
"chromosome",
"biology",
"biology",
"molecular",
"biology",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"yeast",
"and",
"fungal",
"models",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics",
"candida",
"albicans"
] | 2012 | A Coordinated Interdependent Protein Circuitry Stabilizes the Kinetochore Ensemble to Protect CENP-A in the Human Pathogenic Yeast Candida albicans |
In the last several years , a number of studies have shown that spliceosome assembly and splicing catalysis can occur co-transcriptionally . However , it has been unclear which specific transcription factors play key roles in coupling splicing to transcription and the mechanisms through which they act . Here we report the discovery that Gcn5 , which encodes the histone acetyltransferase ( HAT ) activity of the SAGA complex , has genetic interactions with the genes encoding the heterodimeric U2 snRNP proteins Msl1 and Lea1 . These interactions are dependent upon the HAT activity of Gcn5 , suggesting a functional relationship between Gcn5 HAT activity and Msl1/Lea1 function . To understand the relationship between Gcn5 and Msl1/Lea1 , we carried out an analysis of Gcn5's role in co-transcriptional recruitment of Msl1 and Lea1 to pre-mRNA and found that Gcn5 HAT activity is required for co-transcriptional recruitment of the U2 snRNP ( and subsequent snRNP ) components to the branchpoint , while it is not required for U1 recruitment . Although previous studies suggest that transcription elongation can alter co-transcriptional pre-mRNA splicing , we do not observe evidence of defective transcription elongation for these genes in the absence of Gcn5 , while Gcn5-dependent histone acetylation is enriched in the promoter regions . Unexpectedly , we also observe Msl1 enrichment in the promoter region for wild-type cells and cells lacking Gcn5 , indicating that Msl1 recruitment during active transcription can occur independently of its association at the branchpoint region . These results demonstrate a novel role for acetylation by SAGA in co-transcriptional recruitment of the U2 snRNP and recognition of the intron branchpoint .
Eukaryotic genes are interrupted by stretches of noncoding sequence ( introns ) , which are removed from the newly-synthesized RNA by the spliceosome , a dynamic ribonucleoprotein complex made up of 5 highly structured snRNAs and over a hundred snRNA-associated proteins . Although RNA synthesis and RNA splicing have been analyzed as biochemically separate reactions , recent studies demonstrate that these processes are spatially and temporally coordinated [1] . In vivo , recognition of splice sites within the pre-mRNA by the spliceosome can occur while the polymerase is actively engaged with the DNA template [2]–[6] , and recent chromatin IP studies ( in yeast and in mammals ) suggest that this recruitment , or at least the stable association of snRNPs with the transcription complex , occurs in response to synthesis of specific signals in the pre-messenger RNA [7]–[10] . The regulatory implications of this coordination are suggested by studies showing that changes in transcription elongation caused by changes in the activity of specific transcription factors or the presence of transcriptional inhibitors can affect the spliceosome's recognition of splice sites [11] , [12] . These studies focus on the spliceosome's use of alternative splice sites in response to transcription signals , but they raise the possibility that constitutive splicing signals are also affected by conditions or factors that modulate transcription . Despite the evidence that co-transcriptional spliceosome assembly occurs , there is much to learn about the mechanism whereby splicing factors are co-transcriptionally recruited . Transcription of DNA is strongly influenced by its packaging . The core histone proteins , H2A , H2B , H3 , and H4 form an octameric complex that DNA is wrapped around to form the nucleosome , which is further compacted into chromatin—a general repressor of transcription . However , histones undergo extensive post-translational modifications on their N-terminal tails including acetylation , ubiquitination , methylation , and phosphorylation , which alter the chromatin and , in turn , affect transcription . One of the best-characterized histone modifications is the reversible acetylation of lysine residues on the N-terminal tails of histones H2B , H3 , and H4 . Histone acetylation , which is a positive mark of transcription , neutralizes the charge on the basic histone proteins leading to relaxation of the protein/DNA interactions , and the acetylated histone tails can serve as binding sites for proteins that regulate transcription . Histone acetylation is carried out by several different acetyltransferases , the best characterized of which is the protein Gcn5 , a component of the multi-subunit transcription co-activating SAGA ( Spt/Ada/Gcn5/Ada ) complex ( STAGA in mammals ) . Gcn5 primarily acetylates histones H3 and H2B , and these modifications are thought to loosen chromatin for specific transcription factor binding . Furthermore , association between the SAGA complex and general transcription factors , such as TBP , facilitate preinitiation complex formation [13] , [14] . Gcn5 affects global acetylation of histones throughout the genome [15] , but is typically found at the promoter and within coding regions and can influence elongation in addition to events at the promoter [16] . The co-transcriptional nature of pre-messenger RNA splicing raises the intriguing possibility that proteins involved in transcription and histone modification might affect splicing and its regulation . In fact , biochemical studies using mammalian cells indicate that histone-modifying enzymes that regulate histone acetylation physically interact with splicing factors . Prp4K , a U5 snRNP-associated kinase , copurifies with N-CoR , a nuclear hormone corepressor complex that mediates histone deacetylase activity and the mammalian chromatin remodeling protein Brg1 [17] . In an independent affinity purification/mass spectrometry analysis , N-CoR was also found associated with SAP130 and SF3a120 , components of the U2 snRNP that stabilize U2 snRNP-branchpoint interactions [18] . Interestingly , SAP130 also copurifies with the human STAGA complex containing hGcn5 [19] . These studies suggest that mammalian complexes that regulate histone acetylation and chromatin remodeling have physical interactions with splicing factors , although the nature of these interactions remains unclear . Based upon the spatial and temporal proximity of chromatin , chromatin-modifying enzymes ( such as Gcn5 ) , and pre-mRNA splicing complexes during gene expression , we undertook an analysis of genetic interactions between GCN5 and genes encoding nonessential splicing factors . Here we show that deletion of the gene encoding Gcn5 ( and not other yeast lysine acetyltransferases that target histones ) is synthetically lethal when combined with deletion of either gene encoding the U2 snRNP proteins Lea1 and Msl1 ( mammalian U2A′/U2B″ ) . A mutation in GCN5 that eliminates the protein's catalytic activity is sufficient to confer the synthetic lethality . Co-transcriptional recruitment of the U2 snRNP to the branchpoint and subsequent steps in spliceosome assembly are dependent on Gcn5 HAT activity . While previous studies indicate that transcription elongation can alter co-transcriptional spliceosome assembly , chromatin IP results reveal no obvious changes in elongation in the absence of Gcn5's HAT activity . Moreover , we observe a dramatic peak in Gcn5-dependent acetylation of histone H3 in the promoters of these intron-containing genes . Unexpectedly , we also find recruitment of Msl1 at the promoter region , indicating that Msl1 recruitment during active transcription can occur independently of its association at the branchpoint region . These results demonstrate a novel role for acetylation by SAGA in co-transcriptional recruitment of the U2 snRNP and recognition of the intron branchpoint .
In order to characterize interactions between the non-essential histone acetyltransferase GCN5 and genes encoding factors involved in pre-mRNA splicing , a targeted genetic screen was performed to identify synthetic lethal interactions between null alleles of non-essential splicing factors and GCN5 . In this analysis , we uncovered genetic interactions between GCN5 and the splicing factors MSL1 and LEA1 ( Figure 1A ) . Msl1 and Lea1 are the yeast homologs of the human U2 snRNP proteins U2A′/B″ and , like their mammalian counterparts , are components of the U2 snRNP that bind to a conserved stem-loop structure in the U2 snRNA ( Stem-loop IV ) [20] . In vitro , spliceosome assembly is blocked prior to addition of the U2 snRNP in the absence of either Lea1 or Msl1 , indicating a role for these proteins in U2 snRNA association with the pre-mRNA . Cells deleted of either gene also have a mild growth defect , which is observable in the strain background used here . To determine if Gcn5's catalytic activity is required for the interactions with the genes encoding Msl1 and Lea1 , we analyzed specific mutants in the HAT domain of Gcn5 . A previously characterized mutation in GCN5 which changes amino acids 126–128 ( KQL ) in domain I to alanines eliminates the histone acetyltransferase activity of Gcn5 [21] . The effect of this allele was tested in the double mutants , and the KQL mutant is unable to support growth of either gcn5Δ msl1Δ or gcn5Δ lea1Δ double mutant ( Figure 1B ) . By contrast , a mutation in the same domain that changes amino acids 120–122 ( LKN ) to alanines and does not affect Gcn5 HAT activity [21] supports growth of the double mutants ( Figure 1B ) . These results demonstrate that the acetyltransferase activity of Gcn5 is critical for the functional interactions with Msl1 and Lea1 . We also tested other factors that have interactions with Msl1 and Lea1 and are involved in branchpoint recognition , including the commitment complex protein Mud2 , and the U2 snRNP proteins Cus2 and Cus1 , and found no genetic interactions between these factors and GCN5 ( Figure 2A ) . These results demonstrate specificity in the interaction between GCN5 and MSL1 or LEA1 . While we cannot exclude the possibility that there are other essential components of the U2 snRNP that interact with GCN5 , the effect is not general for all splicing factors acting at the prespliceosome formation step . In addition to Gcn5 , there are several other HATs that affect gene expression in yeast , including Elp3 , the catalytic component of the elongator complex , and Sas3 , a component of the NuA3 complex . While both histone acetyltransferases share substrates with Gcn5 , and deletion of either gene is synthetically lethal when combined with deletion of GCN5 [22] , neither ELP3 deletion nor SAS3 deletion has a synthetic interaction with LEA1 or MSL1 ( Figure 2B ) , suggesting that the interactions between GCN5 and MSL1 and LEA1 are specific to the activity of Gcn5 and are not a general feature of all histone acetyltransferases . In addition to acetyltransferases , several deacetylases have been shown to act on the same histone residues as Gcn5 . The histone deacetylase Rpd3 regulates transcription and silencing , and has genetic interactions with Gcn5 [23] . Additionally , Hos2 and Hos3 are involved in gene activation and have been shown to deacetylate histones within the body of genes [16] , [24] . Mutation of HOS2 suppresses gcn5Δ elp3Δ phenotypes [22] . When deletion of RPD3 , HOS2 , or HOS3 is combined with deletion of MSL1 or LEA1 , cells grow indistinguishably from either deletion alone ( Figure 2C ) , suggesting that the acetylation activity of Gcn5 is functionally related to the activities of Msl1/Lea1 , while the removal of acetyl groups from histones probably is not . SAGA is a 1 . 8 MDa , multisubunit complex comprised of five domains containing distinct sets of subunits [25] . Interactions between Msl1 and Lea1 and these other components of the complex were also analyzed and are summarized in Table 1 . Ada2 and Ada3 directly interact with Gcn5 , are required for Gcn5 catalytic activity , and direct Gcn5's histone acetylation activity toward nucleosomes [26]–[29] . We hypothesized that , since abrogation of the catalytic activity of GCN5 leads to synthetic lethality in cells deleted of MSL1 and LEA1 , a similar synthetic growth defect would be evident in the ada2Δ msl1Δ or the ada2Δ lea1Δ mutants , and indeed , this is what is observed . Furthermore , deletion of SPT7 , which is required for the structural integrity of the SAGA complex [25] , [30] , [31] , is lethal when combined with deletion of either MSL1 or LEA1 , indicating that the interactions occur within the context of a functional complex . Two components of SAGA that target the complex to the promoter , Spt3 and Spt8 [31]–[33] , also have genetic interactions with Msl1 and Lea1 . Spt8 is unique to the SAGA complex and is missing from the other Gcn5 containing complexes , SALSA and SILK [34] , suggesting that the interactions between GCN5 and MSL1 and LEA1 occur within the context of the SAGA and not the SALSA or SILK complexes . Deletion of genes encoding other components of SAGA that do not contribute to SAGA's HAT activity , such as Ubp8 or Sgf11 , show no synthetic growth defects when combined with deletion of GCN5 . Taken together , these data strongly suggest that the intact SAGA complex , with its catalytic activity targeted to nucleosomes , has a functional interaction with Msl1 and Lea1 . The best-characterized substrates of Gcn5 are lysine residues on histones , suggesting a model in which chromatin modification has some overlapping function with pre-mRNA splicing factors . Nonetheless , we considered the possibility that the genetic interactions we observed between GCN5 and MSL1 and LEA1 are due to Gcn5's catalytic activity being directed toward one of these non-histone substrates . Using an antibody that recognizes acetylated lysine residues we probed an immunoprecipitated Lea1-HA sample and an Msl1-HA sample to detect acetylation of these proteins in the presence or absence of Gcn5 and do not detect acetylation of either protein or associated U2 snRNP proteins ( data not shown ) . While this does not rule out the possibility that Gcn5 acetylates some other splicing factor , these data do suggest that the genetic interactions between GCN5 and MSL1 and LEA1 are probably not due to acetylation of the U2 snRNP proteins by Gcn5 , and indicate a novel functional interaction between the transcriptional co-activator complex , SAGA , and core components of the spliceosome . Recent studies in yeast demonstrate that in vivo spliceosome recruitment to pre-mRNA occurs while the nascent RNA is actively engaged with the transcription complex [8] . Chromatin immunoprecipitation provides a powerful tool for detecting this co-transcriptional recruitment . The individual snRNPs can be formaldehyde crosslinked to the transcription complex or to the nascent RNA and immunoprecipitated . When the associated DNA is amplified , the signal is enriched in regions of the gene where the snRNPs would be predicted to associate , in a stepwise manner , with the corresponding pre-mRNA [8] . To determine if co-transcriptional recruitment of either Msl1 or Lea1 is affected by deletion of GCN5 , we analyzed the well-characterized intron-containing gene DBP2 with an extended exon 2 ( Figure 3A ) . In strains in which GCN5 is present , we detect Lea1 recruitment after synthesis of the pre-mRNA branchpoint sequence ( Figure 3B ) , a result consistent with what has been reported by others [8] . However , when GCN5 is deleted , there is a dramatic decrease in Lea1 association with DBP2 ( Figure 3B ) . RNA polymerase association along DBP2 was also examined , and no significant difference between the levels of RNA polymerase at the 5′ and 3′ ends of DBP2 are apparent when GCN5 was deleted . In fact , the polymerase distribution along the gene remains relatively unchanged for GCN5 deleted cells relative to wild-type cells ( Figure 3C ) . To determine if DBP2 exon 2 length influences co-transcriptional Lea1 recruitment , we tested the recruitment of Lea1 to DBP2 lacking the extension on exon 2 . We find that the GFP extension has only a mild effect on the overall signal strength observed in the presence of GCN5 with the primer sets used here , and recruitment of Lea1 is eliminated when GCN5 is deleted regardless of whether exon 2 is extended ( data not shown ) . Our discovery of an essential requirement for Gcn5's HAT activity in its interaction with Lea1/Msl1 led to the prediction that its HAT activity would also be required for the co-transcriptional recruitment of Lea1 , and this is what is observed . When Gcn5's HAT activity is abrogated by the KQL mutation , no co-transcriptional recruitment of Lea1 is observed , whereas Lea1's association is unaffected by the LKN mutation ( Figure 3D ) . Pol II occupancy is not significantly affected by either mutation ( Figure 3E ) . A somewhat trivial explanation of these results is that GCN5 deletion or elimination of its HAT activity decreases the amount of Lea1 , leading to a decrease in its association with the gene . However , total Lea1 protein levels are unchanged in the absence of Gcn5 . Neither are levels of Msl1 protein altered ( Figure 3F ) . Co-transcriptional recruitment of Msl1 to DBP2 was also examined . As previously described , Msl1 association with DBP2 is also enriched in regions downstream of the branchpoint sequence . This enrichment is abrogated when GCN5 is deleted or when Gcn5 HAT activity is eliminated ( Figure 4B and 4D , respectively ) . Consistent with previous studies , we routinely observe that the fold enrichment of Msl1 near the branchpoint ( primer set 4 ) relative to the nontranscribed control is lower than for Lea1 . Again RNA polymerase II occupancy was not significantly altered in the strain deleted of GCN5 ( Figure 4C ) . To examine the specificity of the enrichment of Msl1 within DBP2 , we examined the recruitment of Msl1 ( and Lea1 ) to a region further upstream of the promoter of DBP2 and find that neither protein is significantly recruited to these regions in the presence or absence of GCN5 ( Figure S1B ) , suggesting that recruitment of Lea1 and Msl1 is transcription dependent . These data demonstrate that co-transcriptional Msl1 and Lea1 recruitment to the branchpoint region of the pre-mRNA is dependent upon GCN5 . To determine if Gcn5 affects splicing of DBP2 , we performed qRT-PCR to determine the ratio of unspliced pre-mRNA to total DBP2 RNA . Using this analysis , we reproducibly detect an approximately two-fold increase in the Precursor/Total RNA ratio in GCN5 deleted cells compared to WT cells ( Figure S2A ) . When the genes encoding the splicing factors Msl1 and Lea1 are deleted , we observe a 10–15 fold increase in Precursor/Total RNA ratio relative to WT ( approximately 5–9% total unspliced ) ( Figure S2B ) . While deletion of GCN5 leads to a moderate increase in intron accumulation when compared to deletion of a bona fide splicing factor , this reproducible increase indicates that splicing of DBP2 is sensitive to the absence of Gcn5 . While it is clear that post-transcriptional splicing can occur [8] , at least under optimal growth conditions , when co-transcriptional splicing is abrogated , it is likely that the additive effect of disrupting co-transcriptional splicing across the genome has important implications for optimal cellular function , particularly under conditions in which optimal splicing of particular genes is required for cell viability . This hypothesis is currently being tested . Interestingly , we consistently observe enrichment of Msl1 upstream of exon 1 , within the promoter region of DBP2 , which is illustrated by the amplification observed with primer set 1 ( Figure 4B , compare to primer set 4 , which depicts peak enrichment within the gene ) . The level of Msl1 in this region is only mildly decreased when GCN5 is deleted or its catalytic activity is eliminated . This result is surprising since it suggests that the protein is associated with the chromatin before synthesis of the appropriate RNA signal and that the crosslinking step has captured branchpoint-independent interactions between Msl1 and the transcription complex . Msl1 , but not Lea1 , has been shown by yeast two-hybrid to interact with Ssl2 , a component of TFIIH , and Tra1 , a SAGA subunit that interacts with acidic activators [35] . Furthermore , Msl1 , but not Lea1 , affinity purifies with TAF4 , a subunit of the TFIID complex [36] . These unique interactions between Msl1 and components of the transcription machinery that are predicted to act at or near the promoter suggest that Msl1 may be recruited early during transcription initiation and could form a bridge between transcription and U2 snRNP recruitment . The finding that Gcn5 HAT activity is required for co-transcriptional recruitment of the U2 snRNP to DBP2 leads to the prediction that acetylation of DBP2-bound histones is also Gcn5 dependent . To test this prediction , ChIP was performed using an antibody that recognizes diacetylated histone H3 . Histone H3 acetylation peaks at the promoter region of DBP2 ( Figure 4F ) with little evidence of enriched acetylation in the body of the gene . This acetylation drops dramatically when GCN5 is deleted , demonstrating that DBP2-bound histones are acetylated in a Gcn5-dependent manner . Since histone deacetylases ( HDACs ) have been shown to affect rapid/dynamic histone acetylation patterns we examined histone acetylation in the absence of the HDACs shown in Figure 2C , namely Rpd3 , Hos2 , and Hos3 . We found that deletion of these HDACs did not significantly affect acetylation at the promoter or in the body of the gene ( data not shown ) . It remains possible that other deacetylases or some combination of HDACs may contribute to regulation of histone marks involved in co-transcriptional splicing . It is also possible that histones are being rapidly exchanged such that the relevant marks within the body of the gene that facilitate co-transcriptional recruitment of Msl1 and Lea1 are difficult to detect . Nonetheless , Gcn5's acetylation activity , most likely toward histones , appears to be a critical determinant of Msl1 and Lea1 recruitment to the branchpoint . The precise role of Gcn5-mediated acetylation of lysine residues on either histone ( H3 , H2B , or H4 ) or non-histone substrates is currently under investigation . Co-transcriptional recruitment of the spliceosome to the emerging pre-mRNA has been shown to occur in a stepwise fashion [8] , [9] . Here we show that deletion of GCN5 severely abrogates the co-transcriptional recruitment of the U2 snRNP . Combined with our genetic analysis , these results strongly suggest a specific role for Gcn5 activity in U2 snRNP function . Nonetheless , it is possible that deletion of GCN5 acts generally to disrupt co-transcriptional recruitment of all snRNPs . To address this , recruitment of a representative component of the U1 snRNP and triple snRNP was examined . Chromatin IP of Prp42 has been shown to be an indicator of U1 snRNP recruitment to intron-containing genes [7] , [8] . To determine if recruitment of the U1 snRNP is altered in the absence of GCN5 , Prp42 association with DBP2 was analyzed . The U1 snRNP associates with the DBP2 pre-mRNA shortly after synthesis of the 5′ splice site , consistent with reports by others ( Figure 5B ) [7] , [8] . Unlike its effect on U2 snRNP recruitment , deletion of GCN5 does not abrogate the recruitment of the U1 snRNP , demonstrating that the U1 snRNP is still being actively recruited to the pre-mRNA in a co-transcriptional manner ( Figure 5B ) . Hence , the observed disruption of co-transcriptional recruitment of the U2 snRNP in the absence of Gcn5's catalytic activity is specific , and GCN5 deletion does not abrogate all early steps in spliceosome assembly . Since co-transcriptional spliceosome assembly occurs in a stepwise fashion , the prediction is that disruption of U2 snRNP recruitment due to deletion of GCN5 would affect co-transcriptional spliceosome assembly downstream of the U2 snRNP . Snu114 is a U5 snRNP protein that is involved in the destabilization of U1 and U4 snRNAs during spliceosome assembly [37]–[39] . Chromatin IP of Snu114 shows that the U5 snRNP is enriched downstream of the 3′ splice site , a result consistent with previous observations ( Figure 5C ) [8] . However , deletion of GCN5 eliminates the co-transcriptional recruitment of U5 snRNP ( Figure 5C ) , indicating that the lack of U2 snRNP recruitment does alter the recruitment of downstream factors and cripples spliceosome assembly . Although this is consistent with the ordered assembly model of co-transcriptional splicing [8] , [9] , we cannot rule out the possibility of an independent effect by Gcn5 on U5 recruitment . DBP2 was chosen for these studies because of its previously-characterized suitability for chromatin IP studies . DBP2's long intron ( ∼1 Kb ) and long first exon ( ∼1 Kb ) allow for resolution of protein association throughout the gene . We wanted to examine a second well-characterized , intron-containing gene to determine if Gcn5's role in co-transcriptional recruitment of Lea1 and Msl1 is more general . ECM33 has previously been described by others to be a gene to which splicing factors , including the U2 snRNP , are co-transcriptionally recruited [8] . Examination of the co-transcriptional recruitment of Msl1 and Lea1 to ECM33 in the presence of Gcn5 revealed that Lea1 and Msl1 recruitment occurred after the formation of the branchpoint ( Figure 6B and 6C , respectively ) , consistent with what we observed with DBP2 . In the absence of GCN5 , recruitment of Lea1 and Msl1 was abolished ( Figure 6B and 6C , respectively ) . A third gene , YRA1 shows a similar Gcn5-dependent pattern of Msl1 and Lea1 recruitment ( data not shown ) . As with DBP2 , when recruitment of Msl1 and Lea1 to a region further upstream of the promoter of ECM33 was examined in the presence and absence of GCN5 , we find that neither protein is significantly recruited to this region , reinforcing the transcription-dependence of their recruitment ( Figure S1D ) . Also similar to DBP2 , deletion of GCN5 leads to an increase in the Precursor/Total RNA ratio when compared to WT cells ( approximately 4–5 fold ) ( Figure S2A ) . We next examined the acetylation pattern of ECM33-bound histones by ChIP . Consistent with what we observed with DBP2 , a strong Gcn5-dependent peak in acetylation was observed at the promoter of ECM33 , ( Figure 6D ) and little change in this pattern was observed when the HDACs were deleted ( data not shown ) . As with DBP2 , Msl1 recruitment peaks after synthesis of the branchpoint of ECM33 ( Figure 6C , primer sets 3–4 ) . Additionally , analysis of the upstream region of ECM33 shows some early association of Msl1 relative to the non-transcribed control ( Figure 6C ) and especially relative to the peak in signal with primer sets 3 and 4 . This early association of Msl1 is most evident using primer set 2 , although the short distance between amplicons 1 and 2 ( around 300 base pairs ) is likely too small to completely resolve . Nonetheless , the early recruitment of Msl1 to ECM33 is less pronounced than what we observe with DBP2 . A possible explanation for this is the transcriptional frequency of the individual genes . For example , DBP2 generates about 4 times the number of mRNA molecules as ECM33 , and the transcriptional frequency is approximately 7 times greater [40] . It is possible that the increase in transcription of DBP2 allows for more recruitment of Msl1 to the promoter . Taken together , these results suggest that Gcn5-dependent co-transcriptional recruitment of Msl1 and Lea1 to the branchpoint is a common feature among intron-containing genes .
In recent years , there has been strong evidence that splicing can occur co-transcriptionally in yeast and in mammals . However , the mechanism by which spliceosome assembly is coordinated with transcription has been difficult to decipher , particularly in yeast . The genetics and ChIP results described above suggest a model in which Gcn5 mediates co-transcriptional spliceosome assembly by affecting histone acetylation . While we have not detected U2 snRNP acetylation , these data do not rule out the possibility that an additional non-histone substrate ( or substrates ) of Gcn5 can affect co-transcriptional spliceosome assembly , which is something that we continue to explore . It would nonetheless be interesting if Gcn5's acetylation activity is targeted toward a non-histone substrate to abrogate co-transcriptional splicing . Since previous studies have shown that Gcn5 can affect transcription elongation [16] , it is possible that Gcn5 effects on transcription elongation could be responsible for its role in co-transcriptional splicing , especially in light of studies that indicate that changes in elongation can influence pre-mRNA splicing [11] , [12] , [41] . Nonetheless , several lines of evidence suggest that it is not a Gcn5 effect on elongation per se that underlies its role in co-transcriptional snRNP recruitment . First , GCN5 deletion does not appear to significantly affect pol II levels throughout DBP2 or ECM33 . Furthermore , the genetic interactions between MSL1 or LEA1 and GCN5 are not observed with the histone acetyltransferase that acts during elongation , ELP3 . In light of these findings we favor a model in which Gcn5-dependent histone acetylation at the promoter facilitates co-transcriptional recruitment of splicing factors to the branchpoint . We think that it is likely that high promoter acetylation facilitates loading of a factor or factors onto elongating RNA polymerases , and these factors then recruit Msl1 and/or Lea1 to the branchpoint . We cannot rule out that direct interactions between Gcn5 and the U2 snRNP may be important for recruitment , particularly since Gcn5 has been shown to associate both at the promoter and within the body of genes . Although our initial studies do not detect a direct association between Gcn5 and either Msl1 or Lea1 , the interactions may be too weak or transient to detect biochemically . The analysis reported here indicates that deletion of Gcn5 leads to a reproducible increase in unspliced RNA relative to WT cells for both of the genes analyzed . While the amount of unspliced message that accumulates in the absence of Gcn5 is modest on a per gene basis , it is likely that the additive effect across the genome of decreased splicing efficiency when co-transcriptional splicing is abrogated is important . A number of studies of splicing in yeast have found , as we do , that some post-transcriptional splicing can occur even when co-transcriptional splicing is eliminated . Co-transcriptional recognition of splice signals is thought to be a means of increasing the efficiency and perhaps the rate of splicing . Hence , it is likely that conditions under which optimal splicing is necessary will be particularly sensitive to changes in co-transcriptional splicing , which we are currently exploring . We are also testing whether this Gcn5-dependence for optimal splicing increases under growth conditions in which the cell's transcription is particularly dependent on SAGA , which is reported to be the case under a variety of stress conditions [42] . Proper splicing is achieved by sequential recognition of the branchpoint by numerous factors , including the branchpoint binding protein ( BBP ) and the U2 snRNA ( with its associated collection of snRNP-specific proteins ) . The exchange of BBP for the U2 snRNA is the first ATP-dependent step in splicing , and splice sites are committed to participate in this first ATP-dependent step when spliceosomal rearrangements lock the U2 snRNA into place [43] . The work described here suggests that branchpoint recognition is a critical step in coordinating splicing with transcription . A recent mammalian study also suggests that branchpoint recognition is closely tied to transcription . This study identified interactions between U2 snRNP components and the H3K4me3 interacting protein Chd1 . Chd1 bridges U2 snRNP association with trimethylated histone H3 , indicating that U2 snRNP recruitment in mammals is closely tied with transcription and specifically with chromatin “marks” of active transcription [44] . Evidence that a transcriptional coactivator that functions at the 5′ end of the gene can influence U2 snRNP recruitment is particularly interesting in light of a recent proposal that the majority of second exons in yeast may be too short to support stable recruitment of the U2 snRNP and , as a consequence , most endogenous yeast gene splicing is completed post-transcriptionally . Our results suggest that the activity of Gcn5 facilitates co-transcriptional recruitment of the U2 snRNP to at least a subset of genes . Furthermore , co-transcriptional U2 snRNP recruitment may even involve recruitment of Msl1 before synthesis of the branchpoint since Msl1 appears to have unique interactions with the transcription machinery . Our data suggest that the commitment to splicing is likely made co-transcriptionally , and Gcn5 facilitates U2 snRNP association with the pre-mRNA to allow a fluid transition to a U2 snRNP poised to participate in post-transcriptional splicing catalysis . Studies of the mammalian counterpart of SAGA suggest that interactions between the complex and the U2 snRNP may be evolutionarily conserved . Martinez et al . reported that a U2 snRNP protein copurified with the human STAGA complex , although the functional significance of this interaction was not clear [19] . Our results help to explain the functional link between the chromatin modifying machinery and pre-mRNA splicing and demonstrate that Gcn5 , likely within the context of the SAGA complex , has a previously undescribed activity in pre-mRNA splicing .
All S . cerevisiae strains used in this study are listed in Table S1 . Strains described in Table S1 are in the BY4743 strain background , with the exception of Lea1-HA and Msl1-HA strains used for ChIP assays , provided by Karla Neugebauer . All strains were propagated according to standard procedures in either rich media ( YPD ) or appropriate selective media . Plasmid shuffling was performed on 5- fluoroorotic acid ( 5-FOA ) plates . Standard methods for mating , sporulation , transformations , and tetrad analysis were used as described in Methods in Yeast Genetics: A Cold Spring Harbor Laboratory Course Manual . The genotype of each viable spore was confirmed by PCR . Plasmids used in this study are listed in Table S2 . For growth analysis , strains containing a wild-type copy of GCN5 on a centromeric pRS316 ( URA3 ) plasmid were selected for plasmid loss on 5-FOA . Strains were then grown overnight in YPD media at 30°C . Cells were diluted to an O . D . 600 of 0 . 1 in 10 ml of YPD , and incubated at 30°C until all strains reached an O . D . 600 of 0 . 35 . A ten-fold serial dilution of each strain was spotted onto YPD plates and incubated 3–5 days at 30°C . Cells were grown to an O . D . 600 of 1 . 0 and lysed using FA-1 Lysis buffer ( 50 mM HEPES-KOH pH 7 . 5 , 140 mM NaCl , 1 mM EDTA pH 8 . 0 , 1% Triton-X , 0 . 1% Deoxycholate , plus protease inhibitors ) and 0 . 5 mm glass beads with 5 minutes of vortexing at 4°C . The supernatant was cleared by centrifugation and protein concentration was determined by Bradford Assay ( Bio-Rad ) . 50 µg of total protein was fractionated by SDS-PAGE electrophoresis and transferred to a nitrocellulose membrane for immunoblotting with 1∶2000 dilution of anti-PGK1 ( Molecular Probes ) and 1∶1000 dilution of anti-HA 12CA5 ( Roche ) , followed by chemiluminescent detection ( Pierce ) . Cells were grown in YPD to an O . D . 600 0 . 5–0 . 7 and then crosslinked for 15 minutes with formaldehyde to a final concentration of 1% . Cells were disrupted with glass beads ( 0 . 5 mm ) for 40 minutes at 4°C and lysates were cleared by centrifugation . To shear chromatin , lysates were sonicated for a total of six minutes at 30% intensity ( 15 seconds on , 15 seconds off , and on ice ) . After sonication , samples were precleared with CL4B Sepharose beads ( Sigma ) . The precleared samples were then used for immunoprecipitation with either 12CA5 ( Roche ) antibody against the HA epitope or 8WG16 ( Covance ) antibody against RNA pol II . After immunoprecipitation , samples were washed and incubated overnight at 65°C to reverse crosslinking , followed by incubation with Proteinase K ( Sigma ) . DNA was purified using a PCR product purification kit ( Qiagen ) and analyzed by real-time PCR . Input DNA was diluted 1∶20 and 1 µl of this was used in a 25 µl reaction volume . For ChIP DNA , samples were diluted 1∶5 and 1 µl of this was used in a 25 µl reaction volume . Reactions consisted of 12 . 5 µl SYBR GREEN Master Mix ( Applied Biosystems ) and 0 . 5 µM Primers . Real time PCR was performed using an ABI7700 ( Applied Biosystems ) . All samples were run in triplicate for each independent experiment . For quantification , standard curves were generated for each primer set , and DNA concentration for each INPUT and ChIP sample was calculated . ChIP values were divided by the INPUT , and these values were divided by the non-transcribed control and expressed as fold accumulation over the non-transcribed control . Reported values are averages of at least three independent experiments , and error bars represent the standard deviation . For ChIP experiments in Figure 4F and Figure 6D , the ChIP protocol described above was used except samples were sonicated for seven minutes at 30% intensity ( 15 seconds on , 15 seconds , off , and on ice ) . Samples were used for immunoprecipitation with either anti-acetylated histone H3 ( Upstate 06-599 ) or anti-histone H3 ( AbCam ab1791 ) overnight at 4°C . For quantification , standard curves were generated for each primer set . DNA concentration for each INPUT and ChIP sample was calculated using these standard curves and normalized to the non-transcribed control VI_R1 . The normalized IP values calculated for acetylated H3 were divided by the normalized IP values calculated for total H3 . These values are expressed as diacetylated H3 over total Histone H3 . Reported values are averages of three independent experiments , and error bars represent the standard deviation . The data in Figure 5 was generated by standard PCR analysis , ethidium bromide staining , and quantification . The reaction volume was 50 µl , with 0 . 75 µl of template for INPUT , and 5 µl of template for ChIP DNA . Primers were used at a final concentration of 1 µM . PCR products were analyzed on a 1 . 75% agarose gel . Results were quantified using ImageQuant software ( Molecular Dynamics ) . Primer sequences are listed in Table S3 , S4 , S5 . Materials and methods for Figure S1 and Figure S2 are provided in Text S1 . | Pre-messenger RNA splicing , the removal of non-coding RNA sequences ( introns ) that interrupt the protein-coding sequence of genes , is required for proper gene expression . While recent studies have revealed that intron recognition begins while the RNA is actively being synthesized by RNA polymerase II , little is known about how the proteins involved in gene transcription and RNA splicing interact to coordinate the two reactions . Here we show that the protein complex SAGA , which allows RNA polymerase II to navigate the three-dimensional structure of packaged DNA by acetylating histone proteins , has an additional role in pre-messenger RNA splicing . Our genetic analysis shows that the SAGA complex has functional interactions with specific components of the splicing machinery . Furthermore , SAGA's acetylation activity , which we find to be targeted toward promoter-bound histones of intron-containing genes , is required for proper recruitment of these components to RNA during active transcription . Our work supports a model whereby SAGA–dependent acetylation facilitates recruitment of the splicing machinery to the pre–mRNA for proper co-transcriptional splicing . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"biology/histone",
"modification",
"molecular",
"biology/rna",
"splicing"
] | 2009 | Acetylation by the Transcriptional Coactivator Gcn5 Plays a Novel Role in Co-Transcriptional Spliceosome Assembly |
Shotgun metagenomic analysis of the human associated microbiome provides a rich set of microbial features for prediction and biomarker discovery in the context of human diseases and health conditions . However , the use of such high-resolution microbial features presents new challenges , and validated computational tools for learning tasks are lacking . Moreover , classification rules have scarcely been validated in independent studies , posing questions about the generality and generalization of disease-predictive models across cohorts . In this paper , we comprehensively assess approaches to metagenomics-based prediction tasks and for quantitative assessment of the strength of potential microbiome-phenotype associations . We develop a computational framework for prediction tasks using quantitative microbiome profiles , including species-level relative abundances and presence of strain-specific markers . A comprehensive meta-analysis , with particular emphasis on generalization across cohorts , was performed in a collection of 2424 publicly available metagenomic samples from eight large-scale studies . Cross-validation revealed good disease-prediction capabilities , which were in general improved by feature selection and use of strain-specific markers instead of species-level taxonomic abundance . In cross-study analysis , models transferred between studies were in some cases less accurate than models tested by within-study cross-validation . Interestingly , the addition of healthy ( control ) samples from other studies to training sets improved disease prediction capabilities . Some microbial species ( most notably Streptococcus anginosus ) seem to characterize general dysbiotic states of the microbiome rather than connections with a specific disease . Our results in modelling features of the “healthy” microbiome can be considered a first step toward defining general microbial dysbiosis . The software framework , microbiome profiles , and metadata for thousands of samples are publicly available at http://segatalab . cibio . unitn . it/tools/metaml .
The human microbiome constitutes the whole set of microbial organisms associated with the human host . It has been shown to be crucial for human health and for the development and maintenance of the immune system and for several metabolic activities [1–3] . Significant effort has been devoted to its characterization in healthy individuals and subjects with a variety of diseases such as inflammatory bowel diseases [4 , 5] , obesity [6 , 7] , and type-2 diabetes [8] . Consequently , the potential use of the microbiome as a diagnostic tool is a promising line of investigation [9] . In addition , even when the findings are not immediately relevant for the clinical setting , identifying associations between the microbiome and specific diseases is essential for follow-up mechanistic studies . Next-generation DNA sequencing technologies permit comprehensive profiling of the microbial communities from human-associated samples , and have now been sufficiently widely employed to enable meta-analysis for discovering patterns common to independent studies . Meta-analysis has been broadly adopted in other genomics applications , such as for analysis of microarray or RNA-seq data , where multiple studies have been performed for a similar purpose including identifying gene expression signatures of specific human cancers . The general objectives of meta-analysis include proposing new classifiers [10] , comparing different classification methods [11] , finding a common transcriptional profile [12] , and evaluating generalization of prediction models across different studies [13] . In genomics , rigorous meta-analyses are crucial both to validate the findings of each single study , and for providing robust models for clinical purposes . The most common and cost-effective approach for microbiome characterization to date targets the 16S rRNA gene as taxonomic marker [14] . Meta-analyses and independent validation of such experimental approach have identified differences in microbiome composition or function by body site , age , and disease state [15–17] , and have been conducted to determine the most effective techniques for disease classification [18] . More recently , shotgun metagenomics [19] provided expanded resolution to the level of microbial species [20–22] and strain [23] , to the fungal and viral kingdoms [24] , and to the level of individual genes across the metagenome [25 , 26] . The decreasing cost of shotgun metagenomics is rapidly increasing the number of available human disease-associated datasets; however , the generalization of resulting prediction models is still unclear . Improved resolution and lower variability of shotgun metagenomics hold the promise to provide improved generalization of microbial signatures over 16S rRNA sequencing [27] . Meta-analyses on specific host characteristics have been performed ( e . g . , with respect to host age [28] ) . The importance of cross-cohort consistency and validation of predictions has also been recognized , with some works assessing the structure of the microbiome in European cohorts [29] and combined European-American cohorts [21 , 30] . Some studies focusing on the link between host conditions and microbiome further provided a validation step with respect to other single investigations [31 , 32] . Although these works provided a first assessment on the transferability of condition-associated microbiome features across cohorts , no systematic assessments have been performed on clinical outcomes using the full archive of shotgun metagenomic data now publicly available , and no convenient software frameworks for doing so are available in the community . In this study we uniformly process 2424 shotgun metagenomic samples from eight studies to assess the independent prediction accuracy of models built on metagenomic data and to compare strategies for practical use of the microbiome as a prediction tool . The software framework and the microbiome profiles for thousands of samples are made publicly available .
We first assessed the prediction power of metagenomic data in linking the gut microbiome with disease states . For such purpose , we considered six available disease-associated metagenomic datasets spanning five diseases: liver cirrhosis [33] , colorectal cancer [34] , inflammatory bowel diseases ( IBD ) [35] , obesity [31] , and type 2 diabetes ( two distinct studies—[37] and [32] ) . Each dataset was analyzed independently using cross-validation ( denoted as CV in Fig 1 ) , which repeatedly uses part of the samples with associated known phenotype for learning the statistical model , and the remainder for validating the predictions ( see Methods ) . The support vector machines ( SVM ) [38] and random forest ( RF ) [39] classifiers were used for this evaluation as they are state-of-the-art approaches and are appropriate for this type of data [18] . We also evaluated Lasso [40] and elastic net ( ENet ) [41] regularized multiple logistic regression . Neural networks [42] and Bayesian logistic regression [43] represent other possible alternatives not evaluated here . Prediction performance was evaluated by the area under the curve ( AUC ) metric , which summarizes true positive and false positive rates and is robust to unequal proportions of each outcome . Using MetaPhlAn2 species abundance [21] as input data produced high accuracy for disease classification ( Fig 2 ) , although prediction performance varied considerably between datasets . The most predictable disease state appears to be liver cirrhosis ( AUC = 0 . 945 , 95% CI: 0 . 909–0 . 981 for the best classifier ) , followed by colorectal cancer ( AUC = 0 . 873 , 95% CI: 0 . 802–0 . 944 ) , and IBD ( AUC = 0 . 890 , 95% CI: 0 . 812–0 . 968 ) . For IBD we considered Crohn and ulcerative colitis patients together due to the low number of cases in the datasets compared to controls ( as general rule at least ten samples per class are required for reliable prediction models ) . Stronger signatures might be found when considering the two conditions separately with adequate sample size , as it has been observed that some bacterial features are specific to Crohn disease only [16] . Confounding factors such as active treatment could of course lead to overestimated prediction capabilities [44 , 45] , but we adopted here the same contrasting approach used in the original works . For the other diseases we achieved lower discrimination capabilities , suggesting less dramatic microbial shifts in the patients . For type 2 diabetes , although the two considered datasets have independently sampled and geographically distinct cohorts , we obtained very similar AUC values for both ( 0 . 744 , 95% CI: 0 . 688–0 . 800 and 0 . 762 , 95% CI: 0 . 651–0 . 873 for T2D and WT2D , respectively ) . Prediction of obesity generated the lowest AUC ( 0 . 655 , 95% CI: 0 . 576–0 . 734 ) . Despite a wide range of classification performances , all investigated datasets showed a substantial level of association between disease and the microbiome ( Fig 2 ) , with AUC values significantly higher than those obtained by the same classifier applied to the same data with shuffled class labels ( p-values ranging from 9 . 9 × 10−3 for obesity to 5 . 6 × 10−7 for cirrhosis , S1 Table ) . Comparing the accuracy of SVM and RF classifiers , RFs exhibited in all cases similar or better results than SVM . In particular , accuracies differed substantially for three datasets: AUC increased from 0 . 809 to 0 . 873 for colorectal , from 0 . 663 to 0 . 744 for T2D ( difference also supported by statistical significance , p-value 0 . 011 , see S1 Fig ) , and from 0 . 664 to 0 . 762 for WT2D . In two cases , slight improvements were verified: AUC increased from 0 . 922 to 0 . 945 for cirrhosis and from 0 . 862 to 0 . 890 for IBD . Methodologically , our results thus suggested the use of RFs for disease prediction from species abundances . We then investigated how feature selection , i . e . , the procedure of selecting a reduced subset of relevant discriminative features , impacts the prediction accuracy . To this end , we used the RF classifier that implicitly embeds a feature selection step during the model generation phase ( see Methods ) . Feature selection produced a slight improvement of the AUC in all the cases when the model was generated on a reduced set of species ( Fig 3 ) . The advantage of this procedure is twofold . In addition to the increased accuracy , it enables biomarker discovery by detecting the ( few ) species that are most useful to discriminate between “healthy” and “diseased” subjects . These most discriminative species may be prioritized when performing follow-up and validation analyses , and the reduced complexity of the model potentially enables additional evaluations on low-throughput assays . However , the best accuracies were obtained with still relatively high numbers of species , i . e . , more than 60 ( S2 Fig ) . This confirms the complexity of microbial ecosystems where the combination of few species is probably not sufficient to characterize the microbiome associated with complex diseases . We then investigated the use of strain-specific markers , as opposed to species-level taxonomic abundance , by applying the same classification and cross-validation methods to strain-specific microbial features generated by MetaPhlAn2 . Given their strain-specificity , adopting markers as features let us also test the hypothesis that complex diseases are associated with the presence of specific strains or subspecies rather than only species-level abundances . Consistent with this hypothesis , better predictions were obtained from markers ( Fig 2 ) than species abundance , with differences that were statistically significant for one dataset ( S1 Fig ) . This was obtained using SVM with linear kernel , which in this context is more practical to use than RF and SVM with more complex kernels due to the very high dimensionality ( ~100K features ) of the data . Focusing on SVM , markers gave statistically significant improvements with respect to species abundances in half of the datasets ( S1 Fig ) . Moreover , RF in combination with feature selection ( RF-FS:Emb ) achieved satisfactory classification results , i . e . , average accuracies were usually worse than SVM but with no statistically significant difference ( S1 Fig ) even using a very limited portion ( <0 . 2% , S2 Fig ) of the investigated markers . The biomarker discovery step here is of particular interest because it permits identification of a limited set of strain-specific markers potentially directly involved in the association with disease . We also considered alternative approaches to feature selection based on Lasso and ENet ( see Methods ) . Applying Lasso or ENet as pure classifiers , which implicitly incorporates the feature selection and classification steps , did not give satisfactory results , with AUC worse than RF or SVM for both species abundance and marker features ( S3 Fig ) . Better accuracies were obtained by using them for feature selection only , followed by RF or SVM classification . However , both Lasso and ENet feature selection in general worsened the performance of RF and SVM without prior feature selection . Finally , ENet worked better than Lasso , although it was associated with more time-consuming tuning of its free parameters on a two-dimensional grid . Feature selection can also be used for biomarker discovery , and several tools have been developed specifically for this task in metagenomics [46–48] . The approach proposed here ( RF with embedded feature selection ) focuses on the set of features with the most discriminating power rather than on strictly statistical assessments [46] or statistical assessment coupled with effect size [47] . The implemented tool automatically plots the most relevant species ( or markers ) with the importance factor ( see Methods ) along with the average relative abundance ( or average presence ) associated with the different considered classes . We observed a reasonable level of overlap between the detected species and markers , as the most discriminative markers tended to represent strains of the most discriminative species . Interestingly , for all the considered datasets ( Fig 4 and S4 Fig ) the importance factor attributed to each species ( or marker ) was not well correlated with its average relative abundance ( or presence ) in the samples ( maximum correlation of 0 . 49 for the T2D dataset , S5 Fig ) . In several cases , we detected relevant species with partial prevalence but highly discriminative potential between “healthy” and “diseased” subjects . For example , Peptostreptococcus stomatis resulted the most discriminative species in the colorectal dataset with an average relative abundance in the samples less than 0 . 15% . In the cirrhosis dataset , the most relevant taxonomic abundances were enriched in diseased patients . The top features were especially related to the Veillonella ( Veillonella spp . , Veillonella dispar , Veillonella parvula , and Veillonella atypica ) and Streptococcus genera ( Streptococcus anginosus and Streptococcus parasanguinis ) in addition to Haemophilus parainfluenzae , which is consistent with findings of the original study [33] . Species belonging to Veillonella and Streptococcus are typical colonizers of the oral cavity , but they are often overgrown in the small intestine in patients affected by liver cirrhosis , thus suggesting the invasion of the gut from the mouth in these patients [33] . Moreover , species such as Veillonella spp . , V . dispar , V . atypica , and S . anginosus were already associated with opportunistic infections [33] . Also the H . parainfluenzae pathogen may arrive to the gut from the oral cavity [33] . In the colorectal dataset we identified five major species: P . stomatis , Fusobacterium nucleatum ( both enriched in diseased patients ) and Streptococcus salivarius ( depleted in diseased subjects ) as found in the original study [34] , in addition to Parvimonas spp . and Parvimonas micra . We then compared the discriminative species across datasets through hierarchical clustering ( S6 Fig ) . We found some species that were distinctive of one disease only as it is the case for P . stomatis , P . micra and Gemella morbillorum in colorectal cancer , multiple Veillonella species in cirrhosis , and , partially , Bifidobacterium bifidum and Lachnospiraceae in IBD . Interestingly , F . nucleatum was highly discriminant both in colorectal cancer and cirrhosis , suggesting the presence of a similar dysbiosis niche for this organism . Overall , the discriminative species for the two diabetes datasets and the obesity dataset had lower weights , consistent with the lower classification performances achieved with them . Moreover , the pattern of discriminative species for these two datasets clustered together ( S6 Fig ) , suggesting similar dysbiotic configurations of the gut microbiome for obesity and type-2 diabetes . Some species were also found in the set of top discriminative features for all the studies , in particular S . salivarius , S . anginosus , V . parvula , Roseburia intestinalis , and Coprococcus comes . These species might thus be biomarkers of general dysbiosis or ecological community stress in non-healthy states , and should be recognized as such in future disease-microbiome association studies . We extended the cross-validation analysis by evaluating the predictability for non-disease based classification problems . Gender discrimination ( S7 Fig , part a ) exhibited in general low classification accuracy with an AUC close or less than 0 . 6 for most of the considered datasets . However , statistically significant discrimination was verified in some cases ( AUC equal to 0 . 662 and 0 . 796 for skin and IBD dataset , respectively , both p < 0 . 05 by permutation test with shuffled labels ) , which may suggest some gender-dependent differences in the human microbiome as highlighted by recent studies [49] . High classification accuracy in body site prediction in the Human Microbiome Project ( HMP ) dataset ( AUC = 0 . 96 ) , is consistent with previously reported large differences in the microbiome composition among different body areas [1] , and provided validation of the proposed tool for multi-class classification problems ( S8 Fig ) . The confusion matrix revealed moderate misclassification between nasal and skin body sites , which may be due to nasal samples being taken from the anterior nares ( external part of the nostrils ) , and thus having relatively similar biochemical characteristics compared to skin samples from the retroauricular crease . The cross-validation studies discussed in previous sections permitted evaluation of the predictability of different disease states from the human microbiome . However , they are not necessarily a good proxy to evaluate the generalization of the prediction model to independent validation samples , a scenario more relevant to a clinical setting but that has been scarcely investigated . Specifically , how do prediction models perform when applied to samples generated in an independent clinical and laboratory study ? We address this question for several problems of increasing complexity ( denoted as CStaV in Fig 1 ) . We first considered the cirrhosis dataset , in which the samples were acquired in two distinct stages named “discovery” and “validation” ( Fig 5 ) . The generalization of the model was evaluated by ( i ) generating the model on the samples of the training ( TR ) stage and ( ii ) applying it on the test ( TS ) stage . For comparison , we also report the cross-validation results obtained on each specific stage . In general , we found that the model was transferred properly from one stage to the other . In fact , RF applied on species abundance produced an AUC value on the discovery stage that was only slightly decreased from 0 . 936 ( for cross-validation ) to 0 . 919 . For the validation stage we actually obtained an increase from 0 . 958 ( for cross-validation ) to 0 . 972 , and the marker-based predictions achieved slightly better but overall consistent values ( Fig 5 ) . Finally , we note that the AUC achieved on each specific stage were in line with the AUC exhibited by cross-validation using the entire set of samples ( 0 . 945 ) . A similar analysis was done on the T2D dataset , in which samples were collected in two different stages ( stageI and stageII , Fig 6 ) . We verified sufficient generalization of the model across the two stages , although we observed a decrease in accuracy relative to cross-validation . AUC for RF on species abundance decreased from a cross-validation value of 0 . 737 ( 0 . 735 for marker presence ) to 0 . 661 ( 0 . 639 ) for stageI , and from 0 . 743 ( 0 . 771 ) to 0 . 686 ( 0 . 672 ) for stageII . In general , the results obtained on the cirrhosis and T2D datasets provide reasonably good generalization of the model when applied across disease stages , i . e . , to independent samples/batches from the same study . This implies that the samples , although associated with different subjects and acquired at different time points , share common characteristics such the population of study , sample collection approach , DNA extraction protocol , sequencing technology , and analysis strategy [19] . Cross-study validation ( denoted as CSV in Fig 1 ) is a more difficult standard of validation than cross-stage validation , in that training and validation are performed in completely independent studies targeting the same disease . We focused on type-2 diabetes , for which two distinct datasets are available ( i . e . , T2D and WT2D ) . The two datasets presented very different population characteristics as T2D targeted Chinese subjects while WT2D enrolled European women . Still , we observed generalization from one study to the other , ( Fig 6 ) , although cohort effects clearly affected the results . For validation on the T2D dataset , the AUC for RF on species abundance decreased from a cross-validation value of 0 . 744 ( 0 . 747 for marker presence ) to 0 . 569 ( 0 . 566 ) when the model was constructed on the WT2D dataset . Similarly , for validation on the WT2D dataset , AUC decreased from a cross-validation value of 0 . 762 ( 0 . 739 ) to 0 . 664 ( 0 . 622 ) . Different results were achieved by transferring the model to WT2D from the two different experimental stages of the T2D dataset . We obtained an AUC of 0 . 585 ( 0 . 595 ) and 0 . 689 ( 0 . 637 ) by transferring the model from T2D_stageI and T2D_stageII , respectively , indicating that T2D_stageII was more similar than T2D_stageI to WT2D . This similarity was consistent with integrative correlation [50] between the feature relative importance scores obtained on the considered stage of T2D and those on WT2D ( S2 Table ) . The features of T2D_stageII were more correlated to WT2D than were T2D_stageI features , in agreement with the prediction accuracies . Cross-study validation of T2D classification was improved by adding gut microbiome samples from the healthy subjects of four other datasets , i . e . , cirrhosis , colorectal , HMP , and IBD , to the training data . While we included all the control groups as “healthy” , there is the potential for health problems among some control subjects . However , it is standard practice in case-control studies to exclude known disease conditions from control groups , so we can assume that , even in the worst case , just a few diseased patients may be included in the controls and these may be mostly due to undiagnosed cases . In this setting we tested the generalization of the model across cohorts ( Fig 7A ) by generating the models on all the available samples apart those associated with the dataset considered for testing , a "leave-one-dataset-out" cross-study validation [51] ( denoted as lodoCSV in Fig 1 ) . Interestingly , we obtained improved discrimination for T2D when control samples from multiple independent studies were added to training sets , with a high cross-validation AUC score in predicting type-2 diabetes on the entire set of samples ( 0 . 837/0 . 806 for species abundance and marker presence using RF , respectively ) . These values were in fact higher than the AUC obtained by merging all the T2D and WT2D samples into a single set and cross-validating them ( 0 . 743/0 . 736 ) . This cross-validation accuracy was reduced when we tested the generalization of the model to the two T2D datasets ( from 0 . 743/0 . 736 to 0 . 655/0 . 653 and 0 . 709/0 . 679 for T2D and WT2D , respectively ) , which confirmed a non-complete generalization of the model across cohorts . Interestingly , such values obtained by including healthy samples from other cohorts were again higher than for models constructed only on the T2D or WT2D datasets ( Fig 6 ) . Thus , including healthy samples from independent cohorts was effective at improving the detection of T2D status . Finally , we evaluated generalization on the healthy samples of the four other datasets ( prediction assessed in terms of overall accuracy–OA , right part of Fig 7A ) . In such cases we verified high accuracy ( i . e . , OA close to 1 for all the considered datasets ) , confirming correct prediction for most of the control samples . Addition of independent healthy samples to training sets was also performed for gender prediction ( S7 Fig , part b ) , also resulting in increased accuracy , although the discrimination capabilities remained generally low . Overall , these results strongly suggest that the inclusion of samples of healthy individuals from unrelated cohorts is beneficial in disease-targeted investigations , especially when the prediction task has to be generalized to new cohorts . We compared the cross-validation accuracies that we obtained ( Fig 3 ) with results reported in the original papers , when available . For cirrhosis , our best AUC value was 0 . 963 , higher than the cross-validation result reported in [33] ( AUC = 0 . 838 ) . Slight improvements were also verified for colorectal cancer ( AUC = 0 . 881 against the 0 . 84 reported in [34] ) . The best AUC for discrimination of IBD patients in the IBD dataset was 0 . 914 , while a similar analysis was not performed in the original paper [35] . For the other datasets ( i . e . , obesity , T2D , and WT2D ) , the original works used a two-step procedure that tends to overstate discrimination accuracy: i ) first a statistical test was applied on the entire set of samples to select the most discriminative features , then ii ) the model was generated on this set of features and the prediction accuracies were estimated directly on the training set or through a cross-validation approach . This approach overestimates accuracy metrics such as AUC because supervised feature selection is applied on the same data used to evaluate the model , a problem referred by the machine learning community as overfitting [52] . When we adopted the same overfitting-prone procedure , our cross-validation accuracy estimates ( especially using marker features ) were higher than the original ones for all datasets ( S9 Fig ) , but as discussed these are overestimations of the actual discriminative power of the models . Conversely , the overfitting-prone method resulted in much worse performance when the model was transferred to different cohorts . For example , the results reported in [32] showed an AUC equal to 0 . 83 when cross-validating on WT2D , which decreased significantly to 0 . 66 when the model was transferred from the T2D dataset . For the same dataset , we estimated an AUC of 0 . 785 ( Fig 3 ) and 0 . 701 ( Fig 6 ) by ( non-overfitted ) cross-validation and cross-study validation , respectively . Non-overfitted models in general exhibit cross-validation accuracies that are lower , but better represent the ultimate goal of generalization of the model to independent cohorts . We stress that the use of a strict , complete cross-validation/cross-study validation approach is necessary in metagenomics . For cross-validation this requires that for each fold , all training steps ( including feature selection , model selection , and model construction ) are applied on a set of samples that are not overlapping with the samples used for model evaluation/testing . This , together with reducing confounding factors such as antibiotic usage , is necessary for non-overfitted and non-overestimated assessment of the prediction capabilities of metagenomic data . We finally tested the hypothesis that the distinction between the “healthy” and disease-associated gut microbiome can be generalized to diseases for which training information is not available . For this purpose , we considered all gut samples from the disease-associated datasets for a total of 903 samples ( Fig 7B ) . Here the class “diseased” included patients affected by the set of disparate diseases discussed above . The cross-validation analysis on the entire set of samples exhibited satisfactory results ( AUC = 0 . 821 for the best model; most discriminative features are reported in S10 Fig , part a , although in this scenario the model may in reality classify each type of disease separately from the others . More interesting are the results of cross-study and cross-disease prediction . In such cases the disease associated with the testing cohort was not present in the datasets used to generate the model . Although the obtained AUC were lower than the disease-specific cross-validation results reported previously in Fig 3 , we still verified in all cases a certain level of generalization of the model . In particular , the AUC varied between 0 . 628 ( for T2D ) and 0 . 872 ( for cirrhosis ) . This represents an intriguing result that can be associated to the task of modelling the features of the “healthy” microbiome for use as a dysbiosis prediction model for syndromes where few or no training samples are available . As expected , several disease-specific species such as G . morbillorum , B . bifidum and P . micra were not among the most discriminative , the diseases with which they are correlated are not in the training set ( S10 Fig , part b ) . Conversely , species discriminative for multiple diseases ( S . salivarius , S . anginosus , V . parvula , R . intestinalis , and C . comes ) are even more relevant here , confirming that these species are associated to a general non-healthy microbiome state rather than to specific host conditions ( especially S . anginosus , which is the most relevant feature for four of the five testing datasets ) . Overall , this suggests that the dysbiosis-associated microbiome is partially distinct from the healthy microbiome regardless of the specific disease under investigation . This also confirms that study-specific confounding factors [44 , 45] are only partially affecting the estimation of the classification performance . These species identified as associated to general microbiome dysbiosis should be considered in future microbiome studies as non-specific responses to dysbiosis rather than as organisms directly involved in the pathogenesis of the disease under study . We uniformly processed shotgun metagenomic microbiome data for 2424 samples from 8 studies of 6 disease types , and used cross-validation , cross-study validation , and cross-disease validation to evaluate the accuracy of candidate methods of predictive modelling of disease states . We make recommendations of best approaches and non-overfitted practices for using the microbiome as a prediction tool and discuss species and strain-level biomarkers we identified for single and combined datasets . While in this manuscript we focused on taxonomic information , metagenomic functional data such as gene or gene-family abundance data [53] can be exploited in a similar way to conduct a more advanced function-based analysis . Future work will be devoted to exploring more advanced machine learning strategies to further improve classification performance . In general , cross-validation revealed good prediction capabilities , however classification results varied considerably between prediction tasks . In some cases , the ability to predict disease in undiagnosed cases may be overestimated due to the presence of confounding factors such as active antibiotics treatment . The influence of confounding factors on human microbiome has been scarcely investigated in the literature [54 , 55] , but recent studies [44 , 45] highlight this problem and question the study design of some works . Cross-study validation involved evaluating the transferability of prediction models between completely independent patient cohorts . We verified generalization across studies , although transferred models were in some cases less accurate than models tested by within-study cross-validation . Interestingly , the inclusion of healthy ( control ) samples from independent cohorts in training sets was effective for improving the transferability of predictions . We emphasize that considering cross-study performance , instead of the more traditional cross-validation approach , is necessary to understanding prediction capabilities from metagenomic data [56] . Furthermore , avoiding overfitting is crucial for transferring models between different cohorts . Finally , we obtained promising results in the ambitious task of modelling the features of the “healthy” microbiome for use as a dysbiosis prediction model for syndromes where few or no training samples are available . Importantly , this setting is not affected by confounding factors on the target dataset since target samples are not used to build the model . The identified biomarkers for the “healthy” versus “dysbiosis”-associated microbiome are also very important for future microbiome studies of new diseases , because if the same biomarkers are appearing as discriminatory they should be regarded as general dysbiotic organisms rather than microbes directly in involved in the disease under investigation . Compared to the considerable amount of work done on learning methods for 16S rRNA studies , our contribution emphasized two strengths unique to the shotgun metagenomic approach . First , we showed that improved performance can be achieved using strain-level genomic features ( i . e . , markers ) that are not available from 16S rRNA studies . This is also a confirmation that many disease phenotypes are likely linked with microbial genes and factors that are not “core” components of microbial species , but rather are encoded in variable genomic portions that are strain- or subspecies-specific . Second , despite the common potential biases in DNA extraction , shotgun sequencing is considered more consistent across studies than 16S rRNA sequencing for which different variable regions and primer choices are available [57–59] , and thus quantitative microbial signature are inherently less difficult to transfer across cohorts and populations . From a more technical viewpoint , learning analysis in shotgun metagenomes presents distinct challenges due to the very high dimensionality of the dataset when considering strain-level markers ( ~100K features ) , requiring different considerations in machine learning than for 16S rRNA datasets . Altogether , we provide the first validated toolbox for disease prediction across studies using shotgun metagenomics . This study provides a publicly available software framework and uniformly processed microbiome profiles for thousands of samples , to facilitate follow-up studies and evaluation of new methods for classification of disease and other states using metagenomic data . This tool allowed us to assess the predictive power of the microbiome features with respect to disease states and transferability across independent datasets . On a final note , we notice that meta-analyses like the one we performed here were recently regarded as “research parasitism” [60] , because we analyse data produced by other laboratories . The analysis of cross-study predictions , and identification of a dysbiotic microbiome , would not be possible any other way . We hope that not only will these results inform future clinical microbiome studies of disease , but that they promote data transparency and re-use as key components of scientific progress .
We developed a computational tool for metagenomics-based prediction tasks based on machine learning classifiers ( i . e . , support vector machines ( SVMs ) , random forests ( RFs ) , Lasso , and Elastic Net ( ENet ) ) . The tool uses as features quantitative microbiome profiles including species-level relative abundances and presence of strain-specific markers . The framework is fully automatic , including model and feature selection , permitting a systematic and non-overfitted analysis of large metagenomic datasets . Two main kinds of analysis are implemented , i . e . , cross-validation ( to evaluate the prediction strength of metagenomic data ) and cross-study ( to evaluate the generalization of the model between different studies ) . Additionally , the most relevant features are detected for biomarker discovery tasks . Finally , a set of tools is provided to evaluate classification performances in different ways including i ) evaluation metrics such as overall accuracy ( OA ) , precision , recall , F1 , and area under the curve ( AUC ) ; ii ) receiver operating characteristic ( ROC ) curve plots; iii ) confusion matrices; iv ) plots of the most relevant features in addition to average relative abundances; and v ) heatmap figures . The MetAML ( Metagenomic prediction Analysis based on Machine Learning ) tool is open-source and available online at http://segatalab . cibio . unitn . it/tools/metaml . All the species-level taxonomic profiling and marker presence and absence data generated by MetaPhlAn2 and used in this paper are available at the same address . The developed tool incorporates four classification approaches ( i . e . SVM , RF , Lasso , and ENet ) which have been extensively applied in many different fields including computational biology and genomics [18] . The classifiers were implemented using the scikit-learn python package [61] . SVMs aim at finding the hyperplane that maximizes the margin between the samples in different classes [38] , a strategy with many theoretical and practical advantages [62] . Although they are intrinsically linear , they can be extended to the non-linear case by mapping data into a higher dimensional feature space by means of a kernel function . In this work , a radial basis function ( RBF ) kernel was considered for classifying species abundances , while a linear kernel was adopted for markers due to the sparsity of marker-based profiling . In both cases , the best regularization parameter C ( both for linear and RBF kernel ) and the width parameter γ ( only for RBF kernel ) were chosen in {2−5 , 2−3 , … , 215} and {2−15 , 2−13 , … , 23} , respectively , using a 5-fold stratified cross-validation approach . In cross-validation , samples are first randomly subdivided into k subsets ( folds ) of equal size . In particular , we use here stratified cross-validation , in which folds are made to preserve the percentage of samples of each class . A single subset is then used for the testing the model , and the remaining k−1 subsets are used for training . The whole process is repeated k times , with each of the k subsets used once as the testing set . Finally , the results on the k testing folds are averaged to produce a single accuracy evaluation . The parameters that maximize the accuracy ( or another metric of choice ) are finally chosen . SVMs are binary classifiers and , in this work , extension to multi-class classification problems was obtained through the one-against-one approach [63] . Moreover , class posterior probabilities of each sample were estimated from the predicted labels in the binary case using the Platt formulation [64] , which , in the multi-class case , was extended as per [65] . RFs are an ensemble learning method which constructs a large number of decision trees at training time and outputs the class that is the mode of the classes of the individual trees [39] . The free parameters of such classifier were set in this work as follows: i ) the number of trees was equal to 500; ii ) the number of features to consider when looking for the best split was equal to the root of the number of original features; iii ) the quality of a split was measured using the gini impurity criterion . Although a better estimation of such parameters may be obtained through cross-validation , no significant variations were verified by empirical evaluation . We note that RFs can intrinsically deal with binary and multi-class classification problems and give estimation of class probabilities . Moreover , they implicitly provides a list of the features sorted in terms of relative importance . Feature importance was computed in our case using the strategy usually referred to as “gini importance” or “mean decrease impurity” [66] . These importance values were exploited to perform an embedded feature selection strategy ( denoted as RF-FS:Emb ) implemented as follows: i ) RF was applied on the whole set of available features; ii ) features were ranked in terms of importance; iii ) RF was re-trained on the top k-th features , by varying k in the set {5 , 10 , 20 , 30 , 40 , 50 , 60 , 70 , 80 , 90 , 100 , 125 , 150 , 175 , 200}; iv ) the number of features that maximized the accuracy was chosen as the optimal number; v ) the final model was generated by training RF on this reduced set of features . Lasso [40] and ENet [41] are generalized linear modelling approaches that incorporate feature selection and regularization to increase prediction accuracy from high-dimensional and collinear predictors . Lasso is based on a multiple logistic regression trained with L1-norm penalized likelihood , while both L1 and L2 norms are penalized in ENet . In this work , we exploited them in two main ways: i ) directly applying Lasso or Enet as pure classifiers by training a regression model on the binary classification problem; and ii ) using Lasso or ENet for feature selection and then applying SVM or RF on the selected features . In both cases , best regularization parameters were estimated using a 5-fold stratified cross-validation approach . For Lasso this implied to chose the alpha parameter in {10−4 , … , 10−0 . 5} with values evenly spaced on a logarithmic scale . For ENet , along with alpha the L1_ratio parameter was chosen in [0 . 1 , 0 . 5 , 0 . 7 , 0 . 9 , 0 . 95 , 0 . 99 , 1 . 0] . Two main kinds of analysis were performed in this work , i . e . , cross-validation and cross-stages/studies . For cross-validation studies , prediction accuracies were assessed by 10-fold cross validation , repeated and averaged on 20 independent runs . We underline that model selection and feature selection are done using only the training set thus avoiding overfitting problems . In the cross-stages/studies case , all the samples of the first stage/study are considered for training and thus used to generate the classification model including the model selection and feature selection steps . The generalization of the model is evaluated by applying it on the samples of the independent stage/study . In all the cases , the results obtained on the original classification problem were compared with those obtained by a random classifier ( denoted in the paper as SVM-Shuffled and RF-Shuffled ) . For such purpose , we applied the same setting after shuffling randomly the labels of all the samples . Several different metrics were taken into account to evaluate classification performances . First , we considered the OA , which is the percentage of correctly predicted samples . From the confusion matrix three main metrics were computed: i ) the precision ( i . e . , the number of correct positive samples divided by the number of samples predicted as positive ) ; ii ) the recall ( i . e . , the number of correct positive samples divided by the total number of positive samples ) ; iii ) the F1 score , which is the harmonic mean of precision and recall , i . e . , F1 = 2* ( precision*recall ) / ( precision+recall ) . These three metrics ( which range in [0 , 1] , where 1 indicates the best case ) can be computed for each class separately . For brevity , we report in the paper only the average values: after calculating the metrics for each class , their average values , weighted by the number of samples per class , are computed . For binary classification problems , class posterior probabilities were used to plot the ROC curve , which represents the true positive rate ( i . e . , the recall ) against the false positive rate ( i . e . , the number of wrong positive samples divided by the total number of non-positive samples ) . From the ROC curve , we computed the widely-used AUC statistic , which can be interpreted as the probability that the classifier ranks a randomly chosen positive sample higher than a randomly chosen negative one , assuming that positive ranks higher than negative . The AUC ranges in [0 . 5 , 1] , where 0 . 5 corresponds to random change . In the comparison among classifiers , prediction accuracy was assessed by 10-fold cross-validation , repeated and averaged on 20 independent runs . The same folds were used for all classifiers , i . e . training and test sets were identical for each classifier . In this way , the difference in performance of two classifiers could be calculated directly as the difference in AUCs ( or any other metric ) within each test fold . Mean difference and standard error were calculated for each 10-fold CV , then averaged across the 20 repetitions for smoothing . 95% confidence intervals on the difference in AUC performance of two classifiers were calculated using the t-distribution with df = 9 , i . e . : 95%CI:120110∑j=120∑i=110 ( AUC1ij−AUC2ij ) ±2 . 26×σj10 where AUC1ij and AUC2ij are the AUC of two classifiers in fold i of repetition j , and σj is the standard deviation of the AUC1ij−AUC2ij across i = 1…10 folds in repetition j . Similarly , p-values were obtained from the t-statistic obtained with mean difference and standard error smoothed over the 20 repetitions: t=110120∑j=120∑i=110 ( AUC1ij−AUC2ij ) 120∑j=120σj10 using two-tailed t-test with df = 9 , noting that the AUC differences were approximately normally distributed . In terms of feature selection , we reported the list of the 25 most important features found by RFs . For each feature , we considered also the relative importance score , which is a real number in the range [0 , 1] with features that sum to 1 . Feature selection is done for each run independently , and we report the average results . We initially considered a total of 2571 publicly available metagenomic samples ( from eight main studies/datasets ) that were reduced to 2424 after pre-processing and curation ( see next sections ) . These are all the human-associated shotgun metagenomic studies with more than 70 samples and read length bigger than 70nt available as of January 2015 . Six studies were devoted to the characterization of the human gut microbiome in presence of different diseases . Cirrhosis included 123 patients affected by liver cirrhosis and 114 healthy controls [33] . Colorectal consisted of a total of 156 samples , 53 of which were affected by colorectal cancer [34] . IBD represented the first available large metagenomic dataset and includes 124 individuals , 25 were affected by inflammatory bowel disease ( IBD ) [35] . Obesity included 123 non-obese and 169 obese individuals [31] . Two distinct studies were instead related to the alteration of the microbiome in subjects with type 2 diabetes ( T2D ) . In the T2D dataset , 170 Chinese T2D patients and 174 non-diabetic controls were present [37] . The WT2D focused on European women and included 53 T2D patients , 49 impaired glucose tolerance individuals and 43 normal glucose tolerance people [32] . Among these six datasets , two of them comprise two independent stages . For cirrhosis , 181 and 56 samples were collected during the so defined discovery and validation phases , respectively . Similarly , for T2D , 145 and 199 samples were acquired during the first ( stageI ) and second ( stageII ) stages , respectively . Additionally , two studies focused on healthy subjects and not strictly related to the gut microbiome were also taken into account . HMP included samples collected from five major body sites ( i . e . , gastrointestinal tract , nasal cavity , oral cavity , skin , and urogenital tract ) . A subset of these samples were described in [1] . Finally , skin was composed by 291 samples acquired from several different skin sites [36] . The entire analysis was done by taking into account two types of features: species-level relative abundances and presence of strain-specific markers . These features were extracted from the metagenomic samples using MetaPhlAn2 [21] with default parameters . Species abundances are real numbers in the range [0 , 1] that sum up to 1 within each sample , while markers assume binary values . Species abundance and marker presence profiles are characterized by very different numbers of features: in the hundreds for species abundance , and hundreds of thousands for markers ( the exact numbers of features for each dataset are detailed in Fig 2 ) . Before applying MetaPhlAn2 the samples were subject to standard pre-processing as described in the SOP of the Human Microbiome Project [1] without however the step of human DNA removal as these publicly available metagenomes were deposited free from human DNA contamination . Additionally , we removed reads with length less than 90 nucleotides . For the IBD and obesity datasets the minimum length was set to 70 and 75 , respectively , as these cohorts were sequenced with shorter read-lengths . Few samples did not pass the minimum length requirement and were thus discarded . The experimental evaluation can be summarized into five main steps: 1 ) cross-validation analysis was done on the six disease-association datasets for evaluating the capabilities of metagenomic data for disease classification; 2 ) cross-stage studies were performed on the cirrhosis and T2D datasets in order to test the generalization of the model on independent collection batches from the same study; 3 ) in terms of T2D , the analysis was extended by taking into account also samples from completely distinct cohorts; 4 ) cross-studies were also done to model the features of the “healthy” gut microbiome for use as a dysbiosis prediction model for syndromes where few or no training samples are available; 5 ) cross-validation and cross-study analysis were applied to deal with different classification problem such as gender and body site discrimination . We note that all the investigated classification problems , excluding the body site discrimination , represented binary classification problems . Moreover , most of the analysis was done in terms of disease classification , in which the objective was to discriminate between “healthy” and “diseased” subjects . The MetAML ( Metagenomic prediction Analysis based on Machine Learning ) software is open-source , written in Python and available online at http://segatalab . cibio . unitn . it/tools/metaml together with all the data used and discussed in this work . | The human microbiome–the entire set of microbial organisms associated with the human host–interacts closely with host immune and metabolic functions and is crucial for human health . Significant advances in the characterization of the microbiome associated with healthy and diseased individuals have been obtained through next-generation DNA sequencing technologies , which permit accurate estimation of microbial communities directly from uncultured human-associated samples ( e . g . , stool ) . In particular , shotgun metagenomics provide data at unprecedented species- and strain- levels of resolution . Several large-scale metagenomic disease-associated datasets are also becoming available , and disease-predictive models built on metagenomic signatures have been proposed . However , the generalization of resulting prediction models on different cohorts and diseases has not been validated . In this paper , we comprehensively assess approaches to metagenomics-based prediction tasks and for quantitative assessment of microbiome-phenotype associations . We consider 2424 samples from eight studies and six different diseases to assess the independent prediction accuracy of models built on shotgun metagenomic data and to compare strategies for practical use of the microbiome as a prediction tool . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] | [
"body",
"weight",
"medicine",
"and",
"health",
"sciences",
"microbiome",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"neuroscience",
"liver",
"diseases",
"group",
"f",
"streptococci",
"organisms",
"metagenomics",
"physiological",
"parameters",
"forecasting",
"mathematics",
"statistics",
"(mathematics)",
"artificial",
"intelligence",
"gastroenterology",
"and",
"hepatology",
"obesity",
"inflammatory",
"bowel",
"disease",
"bacteria",
"bacterial",
"pathogens",
"microbial",
"genomics",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"medical",
"microbiology",
"mathematical",
"and",
"statistical",
"techniques",
"streptococcus",
"microbial",
"pathogens",
"streptococcus",
"anginosus",
"cirrhosis",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"cognitive",
"science",
"statistical",
"methods",
"machine",
"learning"
] | 2016 | Machine Learning Meta-analysis of Large Metagenomic Datasets: Tools and Biological Insights |
Meiotic recombination is required for proper homologous chromosome segregation in plants and other eukaryotes . The eukaryotic RAD51 gene family has seven ancient paralogs with important roles in mitotic and meiotic recombination . Mutations in mammalian RAD51 homologs RAD51C and XRCC3 lead to embryonic lethality . In the model plant Arabidopsis thaliana , RAD51C and XRCC3 homologs are not essential for vegetative development but are each required for somatic and meiotic recombination , but the mechanism of RAD51C and XRCC3 in meiotic recombination is unclear . The non-lethal Arabidopsis rad51c and xrcc3 null mutants provide an opportunity to study their meiotic functions . Here , we show that AtRAD51C and AtXRCC3 are components of the RAD51-dependent meiotic recombination pathway and required for normal AtRAD51 localization on meiotic chromosomes . In addition , AtRAD51C interacts with both AtRAD51 and AtXRCC3 in vitro and in vivo , suggesting that these proteins form a complex ( es ) . Comparison of AtRAD51 foci in meiocytes from atrad51 , atrad51c , and atxrcc3 single , double and triple heterozygous mutants further supports an interaction between AtRAD51C and AtXRCC3 that enhances AtRAD51 localization . Moreover , atrad51c-/+ atxrcc3-/+ double and atrad51-/+ atrad51c-/+ atxrcc3-/+ triple heterozygous mutants have defects in meiotic recombination , suggesting the role of the AtRAD51C-AtXRCC3 complex in meiotic recombination is in part AtRAD51-dependent . Together , our results support a model in which direct interactions between the RAD51C-XRCC3 complex and RAD51 facilitate RAD51 localization on meiotic chromosomes and RAD51-dependent meiotic recombination . Finally , we hypothesize that maintenance of RAD51 function facilitated by the RAD51C-XRCC3 complex could be highly conserved in eukaryotes .
Homologous recombination ( HR ) is important for repairing DNA damage and maintaining genomic stability . Meiotic HR and sister chromatid cohesion are required for maintaining physical associations between homologous chromosomes ( homologs ) and ensuring their accurate segregation . Meiotic HR is initiated by programmed DNA double-strand breaks ( DSBs ) that are catalyzed by SPO11 , a topoisomerase-like protein [1] . The resulting DSB ends are processed by the MRE11- RAD50-NBS1 ( MRN ) protein complexes to generate 3′ single-stranded DNA ( ssDNA ) tails [2 , 3] , which are subsequently protected by replication protein A ( RPA ) [4] . Functional homologs of the E . coli RecA protein , RAD51 and DMC1 [5 , 6] bind to the 3ʹ tails to form nucleoprotein filaments with the help of several proteins identified in multiple species , including Saccharomyces cerevisiae ( Rad52 [7] , Rad54 [8] , Tid1/Rhd54 [9] , Rad55-Rad57 [10] , Swi5-Sfr1 [11] and PCSS complex [12] ) , Arabidopsis thaliana ( RAD51C [13] , XRCC3 [14] , MND1-HOP2[15] and ATR/ATRIP [16] ) , and mammals ( Mnd1-Hop2 [17] and Brca2-Dss1 [18] ) . The nucleoprotein filaments facilitate single-end invasion of a non-sister chromatid , resulting in the formation of a recombination intermediate called a D-loop , which is then processed to ultimately produce either crossovers ( COs ) or non-crossovers ( NCOs ) [19] . In vertebrate animals and plants , the RAD51 gene family is highly conserved with seven members: DMC1 , RAD51 , RAD51B , RAD51C , RAD51D , XRCC2 and XRCC3 [20–23] , which share Walker A and Walker B motifs with over 37 . 5% similarity [24] . In mice , mutations in any of the paralogs , except DMC1 , lead to embryonic lethality following spontaneous DNA damage or errors [25–29] . In the model plant Arabidopsis thaliana , all seven genes are dispensable for vegetative growth [13 , 14 , 24 , 30–33] . However , AtRAD51 , AtRAD51C and AtXRCC3 are required for somatic and meiotic recombination , as well as plant fertility . Mutations in any of these three genes result in a meiotic chromosome fragmentation phenotype [13 , 14 , 24 , 30–32] . Moreover , AtDMC1 is specifically required for meiotic homolog pairing and recombination [34 , 35] . In contrast to atrad51 , atrad51c and atxrcc3 mutants , atdmc1 mutants do not suffer meiotic chromosome fragmentation; instead their DSBs are thought to be repaired using sister chromatids as templates [34 , 35] . The three other paralogs , AtRAD51B , AtRAD51D and AtXRCC2 , seem to be unnecessary for meiotic DSB repair , because the triple mutant has normal chromosome morphology and fertility [33] . Except for slight differences in synapsis , the chromosome morphology using light microscopy for DAPI-stained chromosomes and fertility phenotypes of atrad51c and atxrcc3 mutants are similar to those of atrad51 , suggesting that their functions are related , but further analyses are needed to understand their mechanistic roles in meiotic DSB repair . Biochemical studies in human cells demonstrate that RAD51 paralogs associate with one another in two distinct complexes: RAD51B-RAD51C-RAD51D-XRCC2 ( BCDX2 ) and RAD51C-XRCC3 ( CX3 ) [36 , 37] . The CX3 complex stabilizes RAD51 binding to ssDNA [36–39] in vitro , thus promoting single-end invasion . Moreover , RAD51C and XRCC3 also help mediate Holliday junction ( HJ ) resolution in vitro [40] , suggesting a later role in meiotic recombination . A yeast two-hybrid assay demonstrated that the Arabidopsis RAD51 paralogs also interact with each other [41] , supporting the idea that RAD51 paralogs function by formation of distinct protein complexes in both animals and plants . However , whether the RAD51 paralogs associate with each other in planta has not been tested . In this study , we report that Arabidopsis homologs of RAD51 , RAD51C and XRCC3 show highly similar meiotic chromosome morphological defects using immune-localization for key markers . We also provide evidence that AtRAD51C and AtXRCC3 are required for AtRAD51 localization on chromosomes . Both in vitro and in vivo data demonstrate that AtRAD51C interacts with AtRAD51 and AtXRCC3 . Furthermore , observation of AtRAD51 foci in atrad51 , atrad51c and atxrcc3 single , double and triple heterozygotes reveals that AtRAD51C and AtXRCC3 both are involved in AtRAD51 loading . Triple heterozygotes also experience non-homolog chromosome associations and have reduced CO frequencies . Together , these results demonstrate that AtRAD51C , AtXRCC3 and AtRAD51 form a complex in planta and are required for AtRAD51 loading on chromosomes .
Previous studies have found that AtRAD51 , AtRAD51C and AtXRCC3 are required for meiotic DSB repair and plant fertility and mutation of individual genes cause indistinguishable chromosome entanglement and fragmentation phenotypes [13 , 14 , 31 , 32] . The similarity of the phenotypes suggests that these genes might function in the same genetic pathway or process . To test this hypothesis , we generated double mutants between atrad51-3 ( SAIL_873_C08 ) [42] , atrad51c ( SALK_021960 ) [13] , and atxrcc3 ( SALK_045564 ) [14] and found that the chromosome morphologies of atrad51 atrad51c ( 48 cells ) , atrad51 atxrcc3 ( 65 cells ) , and atrad51c atxrcc3 ( 54 cells ) double mutants showed no obvious differences compared with each of the single mutants ( S1 Fig ) . The lack of an additive phenotype in the double mutants further supports the hypothesis that they act together in the same biological process . To search for subtle chromosomal phenotypes that could discriminate between the three mutants , we used FISH with a centromere probe for atrad51 ( 82 cells ) ; atrad51c ( 96 cells ) and atxrcc3 ( 81cells ) and a bacterial artificial chromosome ( BAC-F19K16 ) probe that targets a telomere proximal region on chromosome 1 for atrad51 ( 31 cells ) ; atrad51c ( 45 cells ) and atxrcc3 ( 22 cells ) [43] . Wild-type ( WT ) meiocytes had three to five centromere signals at pachytene , indicative of paired homologous centromeres in a cluster ( Fig 1A ) . Although the three mutants had no typical pachytene chromosomes , they all displayed similar centromere clusters or numbers of signals at a stage similar to that of WT , suggesting that AtRAD51 , AtRAD51C and AtXRCC3 are not required for early centromere pairing or clustering ( Fig 1D , 1G and 1J ) . At diakinesis and metaphase I , WT meiocytes had five bivalents , each with two paired centromere signals corresponding to the associated homologs ( Fig 1B ) . In contrast , the three mutants each had 10 centromere signals located on abnormally associated chromosomes ( multivalents-with more than two chromosomes ) ( Fig 1E , 1H and 1K ) , indicating a failure to maintain homolog association , at least at the centromere regions . We next examined homolog pairing on the chromosome arms using the telomere-proximal BAC probe . Unlike the single focus observed on WT pachytene chromosomes , indicative of fully synapsed homologs , meiocytes from each of the three mutants showed two separate signals , indicating a failure to pair properly ( Fig 1M–1P ) . We also performed ASY1 and ZYP1 immuno-localization in WT and mutants . No obvious difference of ASY1 signals at zygotene was found between WT and mutants ( S2 Fig ) . However , unlike WT with linear ZYP1 distribution on pachytene chromosomes , ZYP1 was completely disappeared in rad51 , while some punctate or discontinuous ZYP1 signals were observed in xrcc3 and rad51c ( S2 Fig ) . Together , these results demonstrate that AtRAD51 , AtRAD51C and AtXRCC3 are not required for recombination-independent centromere clustering , but are necessary for homolog pairing , consistent with previous findings obtained using FISH experiment [44] . The similarities of the mutant phenotypes further support the idea that they act in the same process . Loading of RAD51 on ssDNA is aided by several proteins , including Rad52 [45] , Rad55-57 ( Rad51 paralogs ) [46] and Sfr1-Swi5 [11] in yeast , the Brca2-Dss1 complex in mammalian cells [47] , and also by AtBRCA2 in Arabidopsis [48] . The similarity of meiotic defects in Arabidopsis rad51 , rad51c and xrcc3 mutants suggests the RAD51 paralogs RAD51C and XRCC3 may function in meiotic recombination by affecting RAD51 function . To test this hypothesis , we performed an immunofluorescence assay using an AtRAD51 antibody [49] . In Arabidopsis , formation of DSBs is thought to occur at leptotene [50] . At a similar stage , we found that WT plants had 187 . 7±24 . 5 AtRAD51 foci per meiocyte ( n = 14 ) , but the number of foci was greatly reduced in atrad51c ( 36 . 1±9 . 7 , n = 17; P = 1 . 5E-13 ) and atxrcc3 ( 33 . 7±10 . 3 , n = 34; P = 5 . 7E-13 ) mutant meiocytes ( Fig 2A , 2C , 2D and 2Q ) . In contrast , a parallel experiment did not detect any AtRAD51 foci in atrad51 mutant meiocytes at zygotene ( Fig 2B ) . A similar pattern was also observed using pachytene meiocytes ( Fig 2E–2H ) . These results provide evidence that Arabidopsis RAD51C and XRCC3 are required for formation of wild type level of RAD51 foci on meiotic chromosomes . This is consistent with the previous findings for Rad51 paralogs in yeast [46] . Nevertheless , the reduction of AtRAD51 foci in atrad51c and atxrcc3 homozygous mutants does not preclude the possibility that normal level DSBs are formed in these mutants . To test whether DSB frequency is altered in atrad51c and atxrcc3 mutants , we examined the distribution of a DSB marker , phosphorylated histone H2AX ( γ-H2AX ) [51] . At zygotene , after DSBs have been formed , no significant differences in the number of γ-H2AX foci were detected between WT ( 189 . 3±26 . 5 , n = 39 ) , atrad51 ( 176 . 7±15 . 5 , n = 19; P = 0 . 062 ) , atrad51c ( 183 . 6±18 . 0 , n = 18; P = 0 . 41 ) and atxrcc3 ( 178 . 3±13 . 5 , n = 19; P = 0 . 097 ) mutants ( Fig 2I–2L and 2R ) . In Arabidopsis , most meiotic DSBs are thought to be repaired during zygotene-pachytene . We found that γ-H2AX foci were obviously reduced in WT ( 56 . 9±15 . 2 , n = 55 ) pachytene meiocytes compared those of atrad51 ( 132 . 1±15 . 4 , n = 13; P = 1 . 5E-11 ) , atrad51c ( 120 . 6±16 . 6 , n = 14; P = 7 . 2E-11 ) and atxrcc3 ( 122 . 2±18 . 8 , n = 18; P = 1 . 7E-11 ) mutants ( Fig 2M–2P and 2S ) . The presence of normal numbers of zygotene γ-H2AX foci and reduced AtRAD51 foci suggests that AtRAD51C and AtXRCC3 are not required for meiotic DSB formation , but are necessary for AtRAD51 loading . In yeast , the Rad51 paralogs Rad55 and Rad57 form a heterodimeric complex to stimulate RAD51 activity [10] . Vertebrate Rad51 paralogs interact with one another to form two distinct complexes: BCDX2 and CX3 [52] . Like vertebrates , Arabidopsis has seven RAD51 paralogs , and previous yeast two-hybrid assays have shown that XRCC3 interacts with both RAD51C and RAD51 [41] . However , whether these proteins interact in planta has not been investigated . As an initial test for potential interactions we used a yeast two-hybrid assay ( Y2H ) and found that AtXRCC3 interacts with both AtRAD51 and AtRAD51C ( Fig 3A ) , consistent with the previously identified interactions in Y2H system [41] . The interaction between AtRAD51C and AtXRCC3 was further supported by a pull-down assay using recombinant fusion protein of glutathione S-transferase ( GST ) with AtRAD51C and an AtXRCC3-His tag fusion protein ( Fig 3B ) . In addition to the previously identified interactions , we also found that GST-AtRAD51 interacts with AtRAD51C-His ( Fig 3B ) . To explore whether these associations also occurred in planta , we examined the interactions using bimolecular fluorescence complementation ( BiFC ) in tobacco ( Nicotiana benthamiana ) cells . Strong nuclear signals , indicating interaction , were observed for AtRAD51C with either AtRAD51 or AtXRCC3 ( Fig 3C ) . These results provide the first direct evidence that plant RAD51 paralogs RAD51C and XRCC3 interact directly with RAD51 in vitro and in planta . A recent study identified a weak atrad51 allele , atrad51-2 [42] , with a T-DNA insertion in the 3′-untranslated region ( UTR ) that results in reduced AtRAD51 protein levels . This mutant had mild chromosome fragmentation and partial synapsis , as well as some bivalent formation with homologs and non-homologs [42] . In contrast , the atrad51-1 null mutant had severe chromosome fragmentation and formed multivalents during meiotic prophase I [31] . These findings suggest that reducing AtRAD51 level might be a strategy for investigating its meiotic function . Alternatively , analysis of double heterozygous mutants in genes encoding components of a complex can reveal phenotypic defects , even though the corresponding single heterozygotes are phenotypically normal [53 , 54] . We hypothesized that double/triple heterozygotes of atrad51 , atrad51c and atxrcc3 might reduce , but not abolish , their interactions in a complex and reveal informative meiotic phenotypes To test this hypothesis , we generated atrad51-/+ , atrad51c-/+ and atxrcc3-/+ double and triple heterozygous mutants and compared their meiotic phenotypes with WT . Analysis of meiotic chromosome morphology after DAPI staining showed that atrad51-/+ , atrad51c-/+ and atxrcc3-/+ single heterozygote meiocytes and atrad51-/+ atrad51c-/+ and atrad51-/+ atxrcc3-/+ double heterozygotes had similar phenotypes compared to WT ( Fig 4A–4L; S3 Fig ) . In addition , meiocytes from atrad51c-/+ atxrcc3-/+ double heterozygotes had chromosome morphology similar to WT at pachytene ( Fig 4M ) , but at diakinesis , WT formed five bivalents , whereas 32 . 8% ( 20 of 61 , n = 61 ) of the atrad51-/+ atxrcc3-/+ double heterozygote meiocytes had non-homologous chromosome associations ( Fig 4N ) . The cell appears to be able to resolve these associations since equal division of chromosomes was observed at anaphase I and II ( Fig 4O and 4P ) . Meiocytes from atrad51-/+ atrad51c-/+ atxrcc3-/+ triple heterozygotes had a more severe non-homolog association phenotype ( 47 . 8% at diakinesis , 22 of 46 , n = 46 , Fig 4R ) and had unequal chromosome segregation at metaphase II ( 16 . 7% , 2 of 12 , n = 12 , Fig 4T ) . No chromosome fragments were observed in the triple heterozygote , suggesting it is still capable of DSB repair . The results also suggest that RAD51C and XRCC3 are functionally more related to each other than either is to RAD51 . Previous studies showed that T-DNA translocation can cause a similar pattern of chromosome association using light microscopy because the translocated chromosome can associated with two normal chromosomes [55 , 56] . We have verified the T-DNA insertion site by sequencing the junction with flanking genomic DNAs and the results indicated that these mutations are not associated with translocations . To test whether meiotic DSB repair is delayed in the heterozygotes , we performed immunostaining experiments using a γH2AX antibody . As mentioned above , WT meiocytes had 189 . 3±26 . 5 ( n = 39 ) and 56 . 9±15 . 2 ( n = 55 ) γH2AX foci at zygotene and pachytene , respectively ( Fig 5A , 5I , 5Q and 5R and Table 1 ) . All single , double and triple heterozygotes had no obvious differences in the number of γH2AX foci at zygotene , but had significantly more foci at pachytene ( Fig 5A–5R , S1 Table ) . Moreover , the double and triple heterozygotes had more foci at pachytene than the single heterozygotes . There are significantly fewer foci in atrad51-/+ ( 78 . 1±19 . 4 , n = 17 ) , atrad51c-/+ ( 83 . 3±10 . 8 , n = 12 ) and atxrcc3-/+ ( 82 . 0±25 . 9 , n = 24 ) ( S1 Table ) compared to the double mutants atrad51-/+ atrad51c-/+ ( 96 . 3±15 . 4 , n = 30 ) , atrad51-/+atxrcc3-/+ ( 100 . 4±14 . 8 , n = 21 ) and atrad51c-/+atxrcc3-/+ ( 105 . 2±24 . 1 , n = 15 ) , which in turn have significantly fewer foci ( S1 Table ) than the triple atrad51-/+ atrad51c-/+ atxrcc3-/+ ( 113 . 3±14 . 8 , n = 46 ) ( Fig 5J–5P and 5R and Table 1 ) . These data suggest that DSB formation is normal in the heterozygotes , but there is a defect in the progression of DSB repair , and that AtRAD51 , AtRAD51C and AtXRCC3 function in this process . Because AtRAD51C and AtXRCC3 are required for normal AtRAD51 localization , we next examined AtRAD51 localization in heterozygous mutant meiocytes . As described above , WT meiocytes have 187 . 7±24 . 5 ( n = 14 ) AtRAD51 foci at zygotene and 51 . 2±14 . 0 ( n = 65 ) foci at pachytene ( see Fig 6A and 6I for examples and Table 1 ) . In contrast , single , double and triple heterozygous mutant meiocytes have significantly fewer AtRAD51 foci at zygotene ( p<0 . 05; Fig 6A–6H and 6Q ) . At pachytene , the three single mutant heterozygotes show no obvious differences in the number of AtRAD51 foci compared with WT ( Fig 6I–6L and 6R ) , but the double and triple heterozygotes exhibited reduced AtRAD51 foci ( p<0 . 05; Fig 6M–6P and 6R ) . These findings are consistent with the earlier results , suggesting that AtRAD51C and AtXRCC3 play related roles in AtRAD51 loading on chromosomes , likely in a protein complex . As described earlier , the weak atrad51-2 allele is capable of forming bivalents and executing recombination [42] . Similarly , the heterozygous plants analyzed here also completed meiotic recombination to some extent and had partial fertility . To examine CO frequencies in comparison between the various genotypes , we counted the number of chiasmata , the physical manifestation of crossing-over , in WT and mutant meiocytes at both diplotene and metaphase I . On average , WT had 10 . 1±1 . 1 ( n = 52 ) chiasmata per meiocyte and no obvious significant differences were observed in the single heterozygotes: atrad51-/+ with 9 . 6±0 . 7 ( n = 20; P = 0 . 072 ) per meiocyte , atrad51c-/+ with 9 . 6±0 . 7 ( n = 21; P = 0 . 052 ) per meiocyte and atxrcc3-/+ with 9 . 6±0 . 8 ( n = 24; P = 0 . 066 ) per meiocyte . The atrad51-/+ atrad51c-/+ , atrad51-/+ atxrcc3-/+ and atrad51c-/+ atxrcc3-/+ double heterozygotes showed a slight , but statistically significant , reduction of chiasmata with 8 . 4±1 . 2 ( n = 14; P = 6 . 0E-05 ) , 8 . 0±0 . 8 ( n = 10; P = 2 . 9E-06 ) and 7 . 1±1 . 0 ( n = 34; P = 1 . 2E-20 ) per meiocyte , respectively ( Fig 7A–7D and 7K ) . The atrad51-/+ atrad51c-/+ atxrcc3-/+ triple heterozygous mutant also had a significant reduction , with only 6 . 9±1 . 0 ( n = 15; P = 2 . 6E-10 ) chiasmata per meiocyte formed ( Fig 7E and 7K ) . Furthermore , the chiasmata numbers per meiocyte of atrad51c-/+ atxrcc3-/+ double heterozygote ( 7 . 1; P values = 2 . 0E-03 and 1 . 1E-02 , respectively ) and the triple heterozygote ( 6 . 9 , P values = 1 . 7E-03 and 8 . 8E-03 , respectively ) were significantly lower than those of the other two double heterozygotes . Arabidopsis forms two types of COs: interference-sensitive Type I COs that require ZMM proteins like MSH4 and MLH1 [57–59] , and interference-insensitive class II COs that are MUS81-dependent [60 , 61] . To assess the impact of RAD51 and its paralogs on Type I COs , we used an AtMLH1 antibody to visualize AtMLH1 foci , in WT , atrad51-/+ atrad51c-/+ , atrad51-/+ atxrcc3-/+ , atrad51c-/+ atxrcc3-/+and atrad51-/+ atrad51c-/+ atxrcc3-/+ meiocytes at diakinesis [59] . On average , WT meiocytes had 9 . 0±1 . 2 foci ( n = 61 , Fig 7F ) , whereas at similar stages , atrad51-/+ atrad51c-/+ , atrad51-/+ atxrcc3-/+ , atrad51c-/+ atxrcc3-/+ and atrad51-/+ atrad51c-/+ atxrcc3-/+ mutants had 7 . 9±1 . 4 ( n = 40; P = 5 . 8E-05 ) , 7 . 7±1 . 6 ( n = 25; P = 5 . 4E-04 ) , 6 . 4±1 . 3 ( n = 39; P = 5 . 9E-16 ) and 5 . 9±1 . 0 ( n = 16; P = 5 . 5E-12 ) foci , respectively ( Fig 7G–7J and 7L ) . The reduction of AtMLH1 foci in the mutants is consistent with the observed reduction in chiasmata , and supports the idea that Type-I COs are reduced in the mutants . Although the CO number was obviously reduced by ~30% in the atrad51-/+ atrad51c-/+ atxrcc3-/+ triple heterozygote , no univalents were observed , consistent with a mechanism that ensures at least one CO per chromosome [62] . If the COs were distributed among the 5 Arabidopsis bivalents randomly , they would follow the Poisson function P ( k COs per bivalent ) = ( λke-λ ) /k ! where λ is the mean number of COs per bivalent . Using this function , from the analyses of 52 WT and 15 atrad51-/+ atrad51c-/+ atxrcc3-/+ triple heterozygote meiocytes , we would expect to find 36 and 19 univalents in WT and the triple mutant , respectively , but none were observed ( Table 2 ) . To further quantify the remaining COs in atrad51-/+ atrad51c-/+ atxrcc3-/+ , we used a flow cytometry-based assay that measures the segregation of transgenes encoding fluorescent marker proteins expressed using a pollen-specific LAT52 promoter ( FTL markers ) [63 , 64] . The number of viable pollen grains is dramatically reduced in atrad51-/+ atrad51c-/+ atxrcc3-/+ , but it was still feasible to measure CO frequencies using this assay . We crossed atrad51-/+ atrad51c-/+ atxrcc3-/+ with line I2b , which carries two FTL markers ( YFP and DsRed ) on chromosome 2 ( Fig 8A ) . Pollen grains which express both fluorescent proteins have not experienced a crossover between the markers , while those that express only one or the other have . The relative abundance of these two classes can be used to calculate the genetic distance between the two markers [65] . We scored 10 , 092 WT pollen grains and 15 , 460 pollen grains from the triple heterozygote ( Fig 8B and 8C ) . The I2b map distance was 5 . 28±0 . 58 cM in WT and 2 . 87±0 . 33 cM in the triple heterozygote ( Fig 8D–8G ) . The genetic distance between the two fluorescent markers was significantly reduced in atrad51-/+ atrad51c-/+ atxrcc3-/+ ( Z score = 185 . 4 , P value << 0 . 01 ) ( Fig 8G ) , consistent with the reduction in chiasmata described above .
RAD51 family members are conserved across species , from yeast to humans [20] . The budding yeast S . cerevisiae has four RAD51 paralogs ( Rad51 , Dmc1 , Rad55 and Rad57 ) [10] , whereas humans have seven paralogs ( RAD51 , DMC1 , RAD51B , RAD51C , RAD51D , XRCC2 , XRCC3 ) [20] . In yeast , Rad55 interacts with Rad57 to form a stable heterodimer [10] . Similarly , in humans , two complexes are formed by the RAD51 paralogs: the BCDX2 and the CX3 complexes [36 , 37 , 39] . Moreover , a recent study in Caenorhabditis elegans showed that the RAD51 paralogs , RFS-1 and RIP-1 , also exist as a heterodimer and interact with RAD51 [66] . Arabidopsis XRCC3 has been shown to interact with both RAD51 and RAD51C using a yeast two-hybrid assay [41] . We confirmed the yeast two-hybrid result ( Fig 3A ) and demonstrated that the AtRAD51C-AtXRCC3 interaction occurs in planta by using pull-down and BiFC assays ( Fig 3B and 3C ) . It is noteworthy that both pull-down and BiFC assays support an interaction between AtRAD51C with AtRAD51 and AtXRCC3 . Our results strongly support the idea that AtRAD51C is a central factor in complex formation , and is associated with AtRAD51 and AtXRCC3 . These findings are consistent with previous results in human cells that show AtRAD51C associates with two protein complexes [36 , 37 , 39] . Our study is also the first time to show that RAD51 paralogs form a protein complex with RAD51 in plants , supporting the hypothesis that formation of RAD51-paralogs associated protein complexes is highly conserved across eukaryotes , including yeast , humans and plants . Previous studies in yeast showed that RAD51 paralogs are unable to form filaments with ssDNA and do not have a direct role in homology search or single strand invasion [10 , 23] . Nevertheless , studies in different organisms have reported that RAD51 paralogs play important roles in promoting RAD51 function in both mitotic and meiotic HR [46 , 67–69] . For example , the yeast Rad55-Rad57 complex has a role in RAD51-dependent HR [10 , 46] . Similar roles have been found for the C . elegans heterodimer of RAD51 paralogs RFS-1/RIP-1 [66] and the human CX3 complex [36 , 37 , 39] . Due to the lack of direct biochemical data , the role of RAD51 paralogs in meiotic HR in planta is unclear . Studies in the monocot model plant , Oryza sativa ( rice ) , showed that the RAD51 paralogs OsRAD51C and OsXRCC3 are required for meiotic DSB repair and mutations in either result in sterility , chromosome entanglement and fragmentation [70 , 71] . These results are consistent with similar findings in Arabidopsis [13 , 14 , 32] . Immunostaining showed that OsXRCC3 is required for OsRAD51C localization on chromosomes [70] , suggesting the existence of a potential OsRAD51C-OsXRCC3 complex in rice . Additionally , the single-end processing proteins OsCOM1 and OsDMC1 no longer associate with DSB sites in rice osxrcc3 , which suggests that OsXRCC3 , and by extension OsRAD51C , might function upstream of OsRAD51 [70 , 72] . Nevertheless , the relationship between OsRAD51 and its paralogs OsRAD51C and OsXRCC3 remains unclear , because a RAD51 antibody is currently unavailable in rice . In the present study , we showed that Arabidopsis RAD51 foci were obviously reduced in atrad51c and atxrcc3 mutants , consistent to the discovery in rice . Together , these studies , in both rice and Arabidopsis , strengthen the idea that AtRAD51 depends on its paralogs for normal function and that this relationship is highly conserved in eukaryotes . Previous studies showed that RAD51 paralogs have a later role in processing meiotic recombination intermediates [40] . Direct evidence to support the RAD51C-XRCC3 complex having a role in the later meiotic recombination process come from the observation that the RAD51C-XRCC3 complex is associated with HJ resolvase activity . Moreover , RAD51C- and XRCC3-defective hamster cells have reduced resolvase activity and HJ progression [40 , 73] . Similarly , the Arabidopsis RAD51 paralogs AtRAD51B and AtXRCC2 were also reported to affect meiotic recombination in terms of CO number [74] . However , mutations in these paralogs show an increase in meiotic recombination frequency [74] , suggesting that they have roles in meiotic CO formation . In the present study , we found that atrad51c atxrcc3 double heterozygous mutant and the atrad51 atrad51c atxrcc3 triple heterozygous mutant have significantly fewer COs ( Fig 7D , 7E and 7I–7L ) , compared with WT . Given that the reduced number of AtRAD51 foci observed in the double and triple heterozygous mutants , we propose that a diminished capacity to form wild type level of RAD51 foci results in fewer COs in the mutants . The previous finding further supports this idea that a weaker atrad51 allele had fewer chromosome fragments and some univalents , and also formed bivalents between homologs and non-homologs [42] . Therefore , we speculate that AtRAD51 could function in two manners , both dependent on the AtRAD51 paralogs AtRAD51C and AtXRCC3 . Most AtRAD51 foci are required for DNA repair using either homologs or sister chromatids as templates without CO formation , while a small number of AtRAD51 foci might play a role in normal CO formation dependent also on AtDMC1 . Therefore , the AtRAD51C-AtXRCC3 is critical for ensuring wild type number of AtRAD51 foci and COs and facilitating proper homolog recombination and association . Based on our results and previous studies , we propose a model for how AtRAD51C and AtXRCC3 function in conjunction with AtRAD51 in meiotic HR ( Fig 9 ) . Meiotic recombination is initiated by programmed DSBs that are catalyzed by AtSPO11-1 and other proteins . The broken ends are further processed by the MRX protein complex to produce ssDNA tails [2 , 75–77] . In WT , interaction between the AtRAD51C-AtXRCC3 complex and AtRAD51 is proposed to alter the latter’s configuration and facilitates its binding with the ssDNA tails , thus resulting in single end invasion . Consequently , repair of the DSBs yields either COs or NCOs . In the heterozygous mutants , the reduced AtRAD51 level is likely insufficient for supporting the AtDMC1 function , consistent with previous studies in both Arabidopsis and yeast showing that normal DMC1 function in meiosis requires RAD51 [6 , 78] . Thus , with reduced amounts of RAD51 proteins , single end invasion is possibly more promiscuous and targets both homologous and non-homologous templates , resulting in multivalent formation . This aspect of the model is supported by the observation that the triple heterozygous mutant and the weak atrad51 mutant had non-homologous associations and reduced COs . In the homozygous mutants , when AtRAD51 is either completely absent or reduced below a threshold , most or all DSBs are unrepaired , leading to severe chromosome fragmentation and chromosome entanglements . Further investigations are needed to establish the precise AtRAD51 thresholds and how the AtRAD51 paralogs maintain the necessary level of AtRAD51 during the single-end invasion process . In summary , meiotic DSB repair is essential for sexual reproduction in eukaryotes including budding yeast , animals and flowering plants . RAD51 paralogs facilitate the establishment of RAD51 at DSBs and mediate and single end invasion . These functions are also highly conserved in eukaryotes . We propose that facilitation of normal RAD51 function by its paralogs , such as RAD51C and XRCC3 , may be a general mechanism for meiotic DSB repair .
The mutants atrad51-3 ( SAIL_873_C08 ) [42] , atrad51c ( SALK_021960 ) [13] , atxrcc3 ( SALK_045564 ) [14] used in this study were shown previously to be null mutants in the Columbia ( Col-0 ) background . atrad51-/+ atrad51c-/+ and atrad51-/- atrad51c-/- mutants were crossed by atrad51-/+ ( male parent ) and atrad51c-/+ ( female parent ) , atrad51-/+ xrcc3-/+ and atrad51-/- xrcc3-/- mutants were crossed by atrad51-/+ ( male parent ) and xrcc3-/+ ( female parent ) , atrad51c-/+ xrcc3-/+ and atrad51c-/- xrcc3-/- mutants were crossed by atrad51c-/+ ( male parent ) and xrcc3-/+ ( female parent ) . Triple heterozygous mutants were crossed by atrad51c-/+ ( male parent ) and atrad51-/+ atxrcc3-/+ ( female parent ) . Plants were grown at 21°C with 16 h light and 8 h dark . Mutant genotypes were confirmed by PCR using the primers described in S2 Table . A minimum of 10 plants were characterized for each mutant . Chromosome spreads were stained with DAPI and centromere FISH , and immuno-localization experiments were carried out as described previously [79] . Rabbit polyclonal AtRAD51 and γ-H2AX antibodies were used at 1:200 fold dilutions and Alexa Fluor 488 Goat Anti-Rat IgG ( H+L ) secondary antibody ( A-21428 , Invitrogen , Carlsbad , CA , USA ) was used at a 1:1000 fold dilution [80] . Chiasmata distribution statistics were performed following the protocol of Sanchez et al . [81] . BAC DNA extraction ( F19K16 ) and probe labeling were described previously [43] . Images of chromosome spreads were obtained using an Axio Imager A2 microscope ( Zeiss , Heidelberg , Germany ) equipped with a digital camera ( Canon , Tokyo , Japan ) , and processed using Photoshop CS ( Adobe Systems , Mountain View , CA ) . Images were initially captured in black & white and , if necessary , globally false-colored post-capture for visual contrast . AtRAD51 and γ-H2AX foci in WT and mutant lines were counted and statistically analyzed using ImageTool version 3 . 0 software ( University of Texas Health Science Center , San Antonio , USA ) . In mutants that lacked synapsis , we distinguished zygotene from pachytene chromosomes by their relative condensation , with pachytene being more condensed than zygotene chromosomes . To construct the vectors for yeast two-hybrid , pull-down and BiFC assays , full-length AtRAD51 , AtRAD51C and AtXRCC3 cDNA were PCR-amplified using Phanta Super-Fidelity DNA polymerase ( Vazyme Biotech Co . , Ltd , China ) and appropriate primers ( S2 Table ) . For the Y2H assay , full-length AtRAD51 and AtXRCC3 cDNA were purified and ligated into pGADT7 pGBKT7 by NdeI and BamHI double-enzyme digestion , and full-length AtRAD51C cDNA was purified and ligated into pGADT7 and pGBKT7 by NdeI and EcoRI double-enzyme digestion . For the BiFC assay , full-length AtRAD51 and AtXRCC3 cDNA was purified and ligated into pXY103 , pXY104 , pXY105 and pXY106 by BamHI and SalI double-enzyme digestion , and full-length AtRAD51C cDNA was purified and ligated into pXY103 , pXY104 , pXY105 and pXY106 by XbaI and SalI double-enzyme digestion . For the pull down assay , full-length AtRAD51 and AtXRCC3 cDNA was purified and ligated into pET32a and pGEX-6P-1 by BamHI and SalI double-enzyme digestion and full-length AtRAD51C cDNA was purified and ligated into pET32a and pGEX-6P-1 by EcoRI and SalI double-enzyme digestion . All constructs were verified by DNA sequencing . Plasmid vectors were transformed into the Y2H gold yeast strain ( pGBKT7 constructs ) or the Y187 yeast strain ( pGADT7 constructs ) using the LiAc/PEG method . Transformants were mated on YPDA medium for 48 h , and selected on SD/–Trp–Leu plates for 36 h . Transformants were then selected on SD/–His–Ade–Trp–Leu with X-α-Gal and AbA plates to test for positive interactions [82] . AtRAD51 , AtRAD51C and AtXRCC3 were expressed in E . coli using the pGEX6P-1 and pET32a plasmids . The tagged proteins were mixed and incubated for 2 h at 4°C , then pulled down by GST beads for 1 h at 4°C . The protein mixture was confirmed by western blotting with a GST antibody ( AG768 , Beyotime Co . Ltd , China ) or a His-tag antibody ( AH367 , Beyotime Co . Ltd , China ) at 1:100 dilutions , followed by application of an horseradish peroxidase ( HRP ) goat anti-mouse IgG ( H+L ) secondary antibody ( A0216 , Beyotime Co . Ltd , China ) at a 1:2000 dilution . BiFC plasmids ( pXY103/104/105/106-RAD51 , pXY103/104/105/106-RAD51C , pXY103/104/105/106-XRCC3 and pXY103/104/105/106 ) were transformed into Agrobacterium GV3101 cells . Transformants were harvested once the OD600 reached 2 . 0 , and resuspended in MES/MgCl2/acetosyringone solution to a final OD600 of 1 . 0 . Cell suspensions were mixed in 1:1 ratios of various combinations , and young Nicotiana benthamiana leaves were infiltrated . Leaves were excised and visualized using a LSM-710 confocal microscope ( Zeiss ) following 36 h incubation [83] . Open flowers from WT plants or atrad51-/+ atrad51c-/+ atxrcc3-/+plants that were hemizygous for the fluorescent-tagged line ( FTL ) interval I2b and either QRT+/+ or qrt-/+ were collected [64] . The flowers ( 50 or more ) were mixed with 1 mL PBS buffer ( 10 mM CaCl2 , 1 mM KCl , 2 mM MES , 5% w/v sucrose , pH 6 . 5 ) supplemented with 0 . 01% Triton X-100 in a 1 . 5-mL microcentrifuge tube . The mixture was vortexed at maximum speed for 2–3 min and the solution filtered through a 70-μm Falcon® cell strainer ( 352350 , Corning Life Sciences , Tewksbury , MA , USA ) at 450 ×g for 2 min at 4°C . The flow-through was resuspended in a fresh tube with 1 mL PBS buffer at 4°C . Flow cytometry analysis was performed using a Gallios flow cytometer ( Beckman Coulter , Inc . ) . Statistical analysis was performed using Kaluza Analysis 1 . 3 software ( Beckman Coulter , Inc . ) using the two-color analysis methods described previously [65 , 84] . Excel 2016 ( Microsoft , USA ) was used to calculate the mean and standard error of the AtRAD51 foci , γ-H2AX foci , MLH1 foci and the chiasmata numbers of WT and mutants . Data was compared using Student’s t-tests and P values were reported as either exact values or Gaussian approximations . | Meiotic recombination and sister chromatid cohesion are important for maintaining the association between homologous chromosomes and ensuring their accurate segregation . Meiotic recombination starts with a set of programmed DNA double-strand breaks ( DSBs ) , catalyzed by the SPO11 endonuclease . Processing of DSB ends produces 3′ single-stranded DNA tails , which form nucleoprotein filaments with RAD51 and DMC1 , homologs of the prokaryotic RecA protein . The eukaryotic RAD51 gene family has seven ancient paralogs , in addition to RAD51 and DMC1 , the other five members in mammals form two complexes: RAD51B-RAD51C-RAD51D- XRCC2 ( BCDX2 ) and RAD51C-XRCC3 ( CX3 ) . To date , the molecular mechanism of CX3 in animal meiosis remains largely unknown due to the essential roles of these two proteins in embryo development . In Arabidopsis , RAD51C and XRCC3 are required for meiosis and fertility , but their specific mechanisms are unclear . Here we present strong evidence that Arabidopsis RAD51 forms a protein complex with AtRAD51C-AtXRCC3 in vivo . Our data also support the previous hypothesis that CX3 promotes RAD51-denpendet meiotic recombination by affecting its localization on chromosomes . Given that the RAD51 , RAD51C and XRCC3 proteins are highly conserved in plants and vertebrates , the mechanism we present here could be important for the regulation of meiotic recombination in both plants and vertebrate animals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"plant",
"anatomy",
"homologous",
"chromosomes",
"chromosome",
"structure",
"and",
"function",
"centromeres",
"metaphase",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"brassica",
"pollen",
"plant",
"science",
"model",
"organisms",
"experimental",
"organism",
"systems",
"dna",
"plants",
"homologous",
"recombination",
"arabidopsis",
"thaliana",
"research",
"and",
"analysis",
"methods",
"chromosome",
"biology",
"biochemistry",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"nucleic",
"acids",
"phenotypes",
"genetics",
"biology",
"and",
"life",
"sciences",
"dna",
"recombination",
"organisms",
"chromosomes"
] | 2017 | Arabidopsis RAD51, RAD51C and XRCC3 proteins form a complex and facilitate RAD51 localization on chromosomes for meiotic recombination |
Advances in genomic studies have led to significant progress in understanding the epigenetically controlled interplay between chromatin structure and nuclear functions . Epigenetic modifications were shown to play a key role in transcription regulation and genome activity during development and differentiation or in response to the environment . Paradoxically , the molecular mechanisms that regulate the initiation and the maintenance of the spatio-temporal replication program in higher eukaryotes , and in particular their links to epigenetic modifications , still remain elusive . By integrative analysis of the genome-wide distributions of thirteen epigenetic marks in the human cell line K562 , at the 100 kb resolution of corresponding mean replication timing ( MRT ) data , we identify four major groups of chromatin marks with shared features . These states have different MRT , namely from early to late replicating , replication proceeds though a transcriptionally active euchromatin state ( C1 ) , a repressive type of chromatin ( C2 ) associated with polycomb complexes , a silent state ( C3 ) not enriched in any available marks , and a gene poor HP1-associated heterochromatin state ( C4 ) . When mapping these chromatin states inside the megabase-sized U-domains ( U-shaped MRT profile ) covering about 50% of the human genome , we reveal that the associated replication fork polarity gradient corresponds to a directional path across the four chromatin states , from C1 at U-domains borders followed by C2 , C3 and C4 at centers . Analysis of the other genome half is consistent with early and late replication loci occurring in separate compartments , the former correspond to gene-rich , high-GC domains of intermingled chromatin states C1 and C2 , whereas the latter correspond to gene-poor , low-GC domains of alternating chromatin states C3 and C4 or long C4 domains . This new segmentation sheds a new light on the epigenetic regulation of the spatio-temporal replication program in human and provides a framework for further studies in different cell types , in both health and disease .
Understanding the role of chromatin structure and dynamics in the regulation of the nuclear functions including transcription and replication , is a major challenge of current research in genomics and epigenomics [1]–[7] . Since the initial sequencing of complete genomes and more than a decade ago of the human genome [8] , the development of new techniques , in particular chromatin immunoprecipitation ( ChIP ) followed by massive parallel sequencing ( ChIP-seq ) [9] , has enabled genome-wide analysis of many epigenetic modifications such as histone modifications , histone variant incorporation as well as of various DNA-binding proteins [6] . These techniques have been extensively applied to various eukaryotic genomes , from budding yeast [10] , to plants [11] , [12] , worm [13] , fly [14] , [15] , mouse [6] , [16] , [17] and human [6] , [16]–[18] , and have led to significant progress in our understanding of the chromatin landscape and its impact on gene regulation , replication origin specification and cell differentiation . Statistical analyses of these multivariate data sets have shown that this huge combinatorial complexity can be reduced to a surprisingly small number of predominant chromatin states with shared features namely four in Arabidopsis thaliana [19] , five in Caenorhabditis elegans [20] and four [21] or five [22] in Drosophila . To our knowledge , no such a drastic dimensional reduction has been reported in mammalian organisms so far . The application of a multivariate Hidden Markov Model ( HMM ) [23] as well as the implementation of adapted pattern-finding algorithm [24] , have confirmed that distinct epigenetic modifications often exist in well-defined combinations corresponding to different genomic elements like promoters , enhancers , exons , repeated sequences and/or to distinct modes of regulation of gene expression such as actualy transcribed , silenced and poised [23]–[26] . Some recent study [27] of chromatin mark maps across nine different human cell types has ultimately identified fifteen main chromatin types which is a relatively limited number of epigenetic states but probably not the optimal complexity reduction one may achieve in human and more generally in mammalian genomes . The analysis of a wide set of chromatin regulators that add , remove or bind histone modifications reported in Ref . [28] , is a very encouraging step in this direction since six major groups or modules of chromatin regulators were shown to encompass the combinatorial complexity and to be associated with distinct genomic features and chromatin environments . How epigenetic mechanisms and gene expression coordinate with DNA replication has been a long-standing question [1]–[6] . On the contrary to bacteria , yeast and viruses , the genomes of multi-cellular eukaryotes have no clear consensus DNA sequence element associated with replication initiation [29] , [30] . Metazoan genomes duplicate through the coordinated activation of hundreds to thousands of replication origins that can be extremely site-specific or poorly defined with a broad site specification [31] . Indeed more origins are prepared in G1-phase than actively needed in S-phase [32] . Epigenetic mechanisms very likely take part in the spatial and temporal control of origin usage and efficiency in relation with gene expression [32]–[37] . For many years , elucidating the determinants that specify replication origins has been hampered by the very limited number of well established origins in human and more generally in mammals ( a few tens versus a few ten thousands expected ) [4] , [32] , [36] , [38] . Only very recently , nascent DNA strands synthetized at origins were purified by various methods to map replication origins genome-wide in different eukaryotic organisms including Arabidopsis thaliana [39] , Drosophila [40] , mouse [40] , [41] and human [18] , [42]–[47] . Despite some inconsistency or poor concordance between certain of these studies [4] , [48] , some general trends have emerged confirming the correlation of origin specification with transcriptional organization [3] , [4] , [32] . The set of replication origins identified so far are strongly associated with annotated promoters and seem to be enriched in transcription factor binding sites [43] , [44] , [49] and in CpG islands [40] , [41] , [43] . However a significant proportion of origins do not seem to be controlled in the same way as gene transcription since they are in regions void of DNase-I-hypersensitive sites ( DHSs ) and of histone marks found at active promoters [3] , [43] . Interestingly , it has been recently reported that replication origins may contain specific nucleotide sequences . Actually G-rich consensus motifs were shown to be associated with Drosophila , mouse and human origins [40] , [47] , [50] . These analysis have opened new perspectives towards the identification of mechanisms governing origin selection in mammals . The recent blooming of genome-wide mean-replication timing ( MRT ) data in yeast [51] , plants [52] , worm [13] , fly [53] , [54] , mouse [55]–[57] and human [58]–[61] has provided the opportunity to establish links between the spatio-temporal program of replication , transcription and chromatin structure [3]–[6] , [62] . It is now well established that in higher eukaryotes , there is a significant correlation between GC-rich and gene-rich regions replicating early in the S-phase and in between AT-rich and gene poor regions replicating late [55] , [58] , [62] . But recent studies in mammals [56] , [59] and Drosophila [63] , have shown that during differentiation , some genes change expression without change in MRT and vice versa , thereby indicating that transcription is not the only controlling factor and that the epigenetically regulated chromatin structure is likely to be part of the game [3] , [4] , [6] , [62] . In good agreement with previous studies in Drosophila [22] , [63] , genome-wide MRT profiles along mouse and human chromosomes in different cell lines reveal a correlation with epigenetic modifications [64] . Early replicating regions tend to be enriched in open chromatin marks H3K4me1 , H3K4me2 , H3K4me3 , H3K36me3 , H4K20me1 and H3K9 and H3K27 acetylation , whereas late replicating zones are mostly associated with H3K9me2 and to a lesser extent with H3K9me3 [56] , [65] . Importantly , the dynamic changes in MRT observed during development come along with some subnuclear repositioning [56] , [57] , [65]–[69] , early replicating euchromatin domains being generally at the interior of the nucleus whereas late replicating heterochromatic domains are more peripheral or near nucleoli [69]–[73] . Recent experimental studies of long-range chromatin interactions using chromosome conformation capture techniques [65] , [74]–[76] have confirmed that 3D chromatin tertiary structure plays an important role in regulating replication timing . In particular , replicon size , which is dictated by the spacing between active origins , correlates with the length of chromatin loops [37] , [77] , [78] . But as questioned in Refs [76] , [79] , [80] , the dichotomic picture proposed in early studies [65] , [74] , [75] , where early and late replicating loci occur in separated compartments of open and closed chromatin respectively , is somehow too simple as previously questioned in a detailed analysis of replication fork velocity [79] . Identifying the epigenetic chromatin regulators of the spatio-temporal program of DNA replication will be a formidable step towards understanding the so-called replicon and replication foci [71] , [81]–[84] in relation with their transcription counterpart , the transcription factories [71] , . Here we perform principal component analysis ( PCA ) [87] and classical clustering [88] on thirteen epigenetic mark maps in the K562 immature myeloid human cell line ( the results of a similar analysis for the lymphoblastoid GM12878 cell line are reported in the Supplementary Data ) at the resolution 100 kb of corresponding available MRT data , with the perspective of identifying the major types of chromatin states in relation with replication timing during S-phase . For this comparative analysis , we use as a guide the so-called replication U-domains that were shown to cover about half of the human genome for 7 different human cell types including ES , somatic and HeLa cells [80] , [89] . In these megabase-sized domains , the MRT has a characteristic U-shape with early initiation zones at the borders and late replication at centers . Remarkably a significant overlap is observed between these replication U-domains in different cell types and also with the so-called skew N-domains [90]–[92] , where the compositional skew profile accumulated in the germline can be decomposed into a replication-associated linearly decreasing component that shaped as a N [92]–[94] and a step-like transcription associated component that increases in magnitude with transcription and changes sign with gene orientation [92] , [93] , [95]–[97] . From the demonstration that the average replication fork polarity is directly reflected by both the compositional skew and the derivative of the MRT [80] , [98] , [99] , it has been argued that the experimental observation of a MRT derivative that behaves as a N inside replication U-domains is the signature of a progressive inversion of replication fork polarity . These large-scale gradients of replication fork polarity in somatic and germline cells initiate from early initiation zones , also called “master” replication origins [100] , [101] , at U/N-domain borders that were found to be hypersensitive to DNaseI cleavage , to be associated with transcriptional activity and to present a significant enrichment in the insulator-binding proteins CTCF , the hallmarks of localized ( ∼200–300 kb ) open chromatin structure [80] , [101] . The analysis of chromatin interaction HiC [80] and 4C [76] data have revealed that these replication U/N-domains indeed correspond to high-order self-interacting chromatin units . The additional observation of a remarkable gene organization inside these domains with a significant enrichment of expressed genes nearby the bordering “master” replication origins [92] , [102] sheds light on these U/N-domains as regions of highly coordinated regulation of transcription and replication by the chromatin structure . These structural and functional units are conserved in mouse [91] , [92] and are robust to chromosome rearrangements [103] which indicates that they are likely to be a major determinant of genome evolution [104] .
We investigated relationships between the genome-wide distributions of eight histone modifications , one histone variant and four DNA binding proteins in the immature myeloid human cell line K562 ( Materials and Methods ) at the 100 kb resolution of corresponding MRT data [61] , [80] . As a first step , we computed the Spearman correlation coefficient of each mark with each other . We next represented the resulting matrix as a heat map after having reorganized rows and columns with a hierarchical clustering based on the Spearman correlation distance ( Equation 1 , Fig . 1 ) . This preliminary analysis was very promising as regards to the possibility of reducing combinatorial complexity . All the epigenetic marks that are known to be involved in transcription positive regulation , namely H4K20me1 , H3K9me1 , H3K4me3 , H3K27ac , RNAPII , CBX3 , H2AZ , H3K79me2 , H3K36me3 , together with the transcription factors CTCF and Sin3A , form a block in the correlation matrix , meaning that they are all correlated with each other . The maximum correlation is actually obtained between the two active promoter marks H3K4me3 and H3K27ac . As suggested in Refs [27] , [105] , all these active marks are likely to occupy similar regions in the genome . In fact , two lines are clearly apart on the hierarchical clustering dendrogram ( Fig . 1 ) . They correspond to the repressive chromatin marks H3K27me3 and H3K9me3 that are respectively associated with the so-called facultative and constituve heterochromatins [105] , [106] . These two marks are recognized by the chromodomains of polycomb ( Pc ) proteins and heterochromatin protein 1 ( HP1 ) , respectively , components of distinct gene silencing mechanisms which likely explains that they are strongly anticorrelated with each other . While H3K9me3 behaves quite independently with respect to most of the active chromatin marks , H3K27me3 correlates to some of them and especially to H4K20me1 , H3K9me1 and CTCF . When further investigating the correlations between the thirteen considered chromatin marks and the MRT ( Fig . 1 ) , we found , consistently with previous works [56] , [59] , [61] , [64] , [65] , a strong correlation for the transcriptionally active marks with early replication . Some moderate correlation was obtained for the Pc associated repressive marks H3K27me3 which contrasts with the significant anticorrelation observed for the constitutive heterochromatin mark H3K9me3 with late replication . In a second step , to objectively identify the prevalent combinatorial patterns of the thirteen chromatin marks , we performed a PCA [107] to reduce the dimensionality of the data ( Materials and Methods ) . We then concentrated on the first three principal components , which together account for 76% of the total data set variance ( Supplementary Fig . S1 ) . By projecting the 100 kb genomic loci on the ( PC1 , PC2 ) plane ( Fig . 2A ) and the ( PC3 , PC2 ) plane ( Fig . 2B ) , we noticed that four areas contain most of the population . On the ( PC1 , PC2 ) plane , a large area of medium density comes out from a plane of much higher density . As viewed on the ( PC3 , PC2 ) plane , in this very dense plane , loci mainly lie along two straight lines with a very high density of loci concentrated at the intersection of these lines . This led us to use the Clara clustering algorithm [88] , which is very similar to k-means , with the number of clusters fixed to four ( Materials and Methods ) . When labeling each of the four main chromatin states with a color , we obtained four domains in the 3D scatter plot ( Fig . 3A ) that have common boundaries as evidenced on the three orthogonal projections on the planes ( PC1 , PC2 ) ( Fig . 3B ) , ( PC1 , PC3 ) ( Fig . 3C ) and ( PC3 , PC2 ) ( Fig . 3D ) . To improve the quality of our clustering procedure , we filtered out poorly clustered data points that are closer to another cluster than to the one they belong to ( black dots in Fig . 3 ) , where the distance between a data point and a cluster is defined as the mean of the distances of this point to all the points in the cluster . Removing those points is exactly equivalent as removing points with a negative silhouette [108] ( Materials and Methods ) . To determine the number of clusters , we used two statistical criteria ( Materials and Methods ) . Four is the optimal choice according to the within-cluster sum of squares that clearly displays an elbow ( abrupt slowing down of the decay ) at the cluster number equal to four ( Fig . 3B ) . The gap statistic [109] indicates that two or four clusters are good solutions ( Fig . 3C ) . Our choice of four main chromatin states ( Fig . 3A ) can thus be seen as an attempt to test the limits of the classical dichotomic picture [65] , [74] , [75] of two chromatin states , one open ( euchromatin ) and another one closed ( heterochromatin ) ( Supplementary Fig . S2A ) . The four prevalent chromatin states so identified and further labeled C1 , C2 , C3 and C4 , were respectively found in 6572 ( 23 . 8% ) , 5312 ( 19 . 2% ) , 6603 ( 23 . 9% ) and 6758 ( 24 . 4% ) among the 27656 100 kb loci with a defined MRT ( Materials and Methods ) . Indeed , we removed from the analysis the 2411 ( 8 . 7% ) loci that were not properly classified in any chromatin state . More than 90% of the loci in C1 are associated ( positive enrichment ) with the histone modifications H3K36me3 , H3K4me3 , H3K27ac and H3K79me2 , the hallmarks of transcriptionally active chromatin ( Fig . 4 ) [2] , [6] , [105] , as well as of the loci associated with RNA Polymerase II ( Fig . 5 ) and the RPD3-interacting protein SIN3A ( Fig . 5 ) as previously found in active euchromatin in Drosophila [22] . The majority of C1 loci are marked by H3k9me1 loci consistently with the observation of higher H3K9me1 levels in active promoters [105] , and also contains the histone variant H2AZ whose binding level was shown to correlate with gene activity in human [105] ( Fig . 4 ) . C2 is notably associated with the histone modification H3K27me3 ( Fig . 4 ) , hence corresponds to a Polycomb repressed facultative heterochromatin state [105] , [106] . Out of the four main chromatin states , C3 corresponds to 100 kb loci that are not enriched for any available marks . C3 can be compared to the “null” or “black” silent heterochromatin regions previously found in Drosophila [21] , [22] and Arabidopsis [19] as covering a significant portion of the genome . C4 corresponds to the classic HP1-associated heterochromatin state with all of the 6603 C4 100-kb-loci containing the H3K9me3 mark and almost only that repressive mark ( Fig . 4 ) [105] , [106] . Methylation of H3K9 is well known to be implicated in heterochromatin formation and gene silencing [2] . The fact that H3K9me1 is found almost equally in C1 and C2 and not in C4 ( Fig . 4 ) , confirms that this epigenetic modification may also be associated with transcriptional activation [105] . H3K9me3 is found in all C4 100-kb-loci as the probable signature of its ability to anchor the heterochromatin protein HP1 at the origin of the establishment of heterochromatin . But H3K9me3 is not exclusively found in C4 loci; indeed 75% of C1 loci and 50% of C2 loci contain some H3K9me3 marks ( Fig . 4 ) . In the transcriptionally active state C1 , H3K9me3 is present in combination with all active marks which might conduct in the anchoring of the isoform of the HP1 protein [110]–[113] , also called CBX3 ( Fig . 5 ) , which was recently shown to help the splicing of multiexonic genes [114] , [115] . The insulator-binding protein CTCF is known to establish chromatin boundaries to prevent the spreading of heterochromatin into transcriptionally active regions [116] , [117] . Consistent with the idea that CTCF-bound insulators prevent heterochromatin to invade genic regions , we found in good agreement with previous observation in Drosophila [21] , [22] that CTCF is contained in C1 loci and to a slightly less extent in C2 loci ( Fig . 5 ) . Despite the original association of H4K20 methylation with repressive chromatin [2] , H4K20me1 was recently shown to strongly correlate with gene activation [105] . In particular when combined with H3K36me3 and H2BK5me1 , this mark was found at highly expressed exons near human gene 5′-ends [118] . The high level of H4K20me1 found in C1 ( Fig . 4 ) is quite consistent with these observations . However , we observed the same level of H4K20me1 in C2 which is silent . This suggests that this mark is not uniquely linked to transcription activation . Interestingly , recent works have confirmed that PR-Set7 involved in the deposition of H4K20me1 plays an important role in the control of replication origin firing in mammalian cells [119]–[121] . To assess the generality of the four prevalent chromatin states , we ran the same clustering procedure on the lymphoblastoid cell line GM12878 and on a third blood cell line ( Monocyte CD14 , Monocd14ro1746 ) . The same four main chromatin states emerged in the three cell lines ( Supplementary Figs S7 , S9 , S10 , S11 ) . Hence the chromatin organization in four chromatin states is shared by at least several somatic human cell lines . This classification into four main chromatin states of the human genome shows strong similarities with those recently reported in Arabidopsis [19] and Drosophila [21] , [22] suggesting the possible existence of some simple principles of epigenetic compartimentalization of eukaryotic genomes . However , what our study reveals with respect to previous works , is a strong correlation between these chromatin states and MRT ( Fig . 6 ) . C1 , C2 , C3 and C4 actually have significantly different MRT probability distribution functions ( Fig . 6A ) with a clear shift from early to late replicating as evidenced by the cumulative distribution functions ( Fig . 6B ) . By applying a wilcoxon test to each pairs of chromatin states , we did verify that the p-value was infinitesimal . The transcriptionally active euchromatin state C1 replicates early in S phase consistent with previous analysis of open chromatin marks in human and mouse [56] , [59] , [61] , [62] , [64] , [65] . The Pc-repressed facultative heterochromatin state C2 is replicated slightly later in mid-S phase which corroborates the recent finding of an association of H3K27me3 with mid-replicating chromosomal domains in human fibroblast [106] . This rather clear observation contrasts with previous contradictory results concerning the existence of high correlation between late replication and this repressive chromatin mark [65] , [122] . The silent unmarked chromatin state C3 replicates later than C2 but before the HP1-associated heterochromatin state C4 that replicates very late almost at the end of S phase ( Fig . 6 ) . As previously reported in Drosophila [22] , [63] , these results confirm the existence of a strong link between epigenetic chromatin states and MRT in human . They further suggest that the epigenetically controlled chromatin structure has some impact on the normal progression of S-phase . To address the question of the gene content of these four prevalent chromatin states , we used a data set of 23818 genes that are spatially distinct ( Materials and Methods ) . Some of these genes ( 3001 ) were not taken into account in our analysis because their promoter don't belong to any chromatin state . The mean density of the 20817 genes that belong to one of the four chromatin states is 8 . 24 promoters per Mb . The only chromatin state that is highly enriched in gene promoters is the early replicating euchromatin state C1 that harbours 62 . 0% of gene promoters even though it represents about 25% of the total genome coverage by the four chromatin states ( Table 1 and 2 ) . The mid S facultative heterochromatin state C2 also contains a non negligible percentage ( 19 . 6% ) of gene promoters that indeed corresponds to a modest density 7 . 7 promoters/Mb as compared to 19 . 1 promoter/Mb found in C1 . The late replicating unmarked and constitutive heterochromatin states C3 and C4 are genuinely gene deserts with very low gene densities 4 . 1 promoters/Mb and 1 . 8 promoter/Mb respectively . The mean gene length increases gradually from C1 to C4 going from 42 . 5 kb to 133 . 1 kb ( Table 1 ) . This discrepancy in gene length explains why the gene coverage decreases less abruptly than the promoter density , with C1 mainly genic ( 62 . 9% ) , C2 modestly genic ( 49 . 8% ) and C3 ( 39 . 5% ) and C4 ( 29 . 3% ) mostly intergenic . To investigate gene expression in chromatin states , we used a data set of 17872 genes with a valid expression value in K562 ( Materials and Methods ) . Of those genes , 15869 belong to one of the chromatin states . We found that a vast majority of expressed genes with a ( Equation 7 ) are in the early replicating euchromatin state C1 ( Fig . 7B ) , which confirms the link between MRT and expressed gene density previously reported in mammals [55] , [58] , [59] , [61] . As expected , most of the genes in the facultative Pc repressed heterochromatin state C2 are non expressed . Interestingly , we found that the density of non expressed genes in C1 is equivalent to the one in C2 , indicating that it is more the predominance of active genes that characterizes early replicating regions than the absence of repressed genes . This explains why the correlation between MRT and gene expression is stronger if one considers the expressed gene density ( , ) ) than the mean expresssion ( , ) as previously observed in Drosophila [54] . Indeed in C1 the mean gene expression level is lowered by the presence of a non negligible set of non-expressed genes . The few genes in the heterochromatin states C3 and C4 are silent except a minority of them . We assessed gene function on the basis of gene ontology [123] . We analyzed the genes in each chromatin states according to their biological process ( Supplementary Fig . S3 ) , component ( Supplementary Fig . S4 ) and function ( Supplementary Fig . S5 ) using GO SLIM annotation ( Materials and Methods ) . We computed the enrichment p-value using the Hypergeometric distribution and used the odd ratio value to determine if the deviation from expected number of genes for the considered GO terms was an enrichment ( ) or a depletion ( ) . As previously observed for gene expression , these GO terms provide some clear discrimination between genes in the early replicating transcriptionally active euchromatin C1 and genes in the repressed heterochromatin states C2 , C3 and C4 . Genes enriched in C1 are almost systematically depleted in C2 , C3 and C4 , whereas on the opposite , genes that are depleted in C1 are enriched in at least one if not all the heterochromatin states C2 , C3 and C4 . We found C1 to be enriched mainly in housekeeping genes . The highest enrichments were obtained for the following process categories: mRNA processing , translation , ribosome biogenesis , DNA metabolic process , chromosome organization and segregation , cell cycle and cell division and for the corresponding component categories: ribosome , chromosome , nucleolus , nucleoplasm , nuclear envelope , mitochondrion and microtubule organizing center . The highly depleted process categories in C1 correspond to tissue specific genes that are not expressed in the immature myeloid K562 cell line as for example neurological system process , extracellular matrix organization , cell adhesion and cell motility , or that are defficient in these cancer cells like circulating system process [124] , [125] . Along the line of the isochore model [126] , GC-rich and GC-poor regions were shown to match the cytogenic R and G bands and to correlate well with early and late replicating domains in mammals [8] , [127] , [128] . GC-rich regions correspond to regions of very high density of genes including the housekeeping genes and associated CpG islands . This also correspond to regions enriched in short inter-dispersed repetitive DNA elements ( SINEs , Alu ) [8] . In contrast , GC-poor regions are definitely poor in genes , predominantly tissue-specific genes containing rather large introns , but are relatively rich in long inter-disperse repetitive DNA elements ( LINES ) [8] that are significantly more abundant in these regions . Consistently , we found that the early replicating euchromatin state C1 has a GC content distribution shifted to higher values as compared to the unmarked and constitutive heterochromatin states C3 and C4 respectively ( Fig . 8A ) . C1 is definitely GC-rich with an mean value that is significantly higher than the genome average ( ) . On the opposite C3 and C4 are GC-poor with and 36 . 7% , respectively . Surprisingly , the Pc repressed facultative heterochromatin state C2 has a GC content distribution similar to the one obtained for C1 ( Fig . 8 ) with . This means that if a high density of early replicating and highly expressed genes implies a high GC content , the reciprocal is not true . For example , C2 loci corresponding to 18% of the genome are GC-rich ( Fig . 8A ) but gene poor ( Table 1 ) and most of these C2 genes are silenced by Pc proteins . Cytosine DNA methylation is a mediator of gene silencing in repressed heterochromatin regions , while in potentially active open chromatin regions DNA is essentially unmethylated [129] , [130] . Methyl-cytosines being hypermutable , prone to deamination to thymines , CpG o/e ratio ( Materials and Methods ) is commonly used as an estimator of DNA methylation , the higher this ratio , the lower the methylation [101] , [131] . When computing CpG o/e after removing the CpG islands ( CGIs ) that are short unmethylated regions rich in CpG , in the four chromatin states , we found a significant shift of the CpG o/e pdf to smaller values when going from C1 ( ) to C2 ( ) , C3 ( ) and C4 ( ) ( Fig . 8 ) . Thus relative to the genome average value , the early replicating transcriptionally active euchromatin state C1 is clearly hypomethylated . The mid-S repressed facultative heterochromatin state C2 is also , but at a lesser extent , less methylated than the entire genome . As expected the late replicating unmarked and constitutive heterochromatin states C3 and C4 are definitely methylated , the later being significantly more methylated than the entire genome . Thus the differences in CpG o/e ( Fig . 8B ) and MRT ( Fig . 6A ) observed in the four chromatin states C1 , C2 , C3 and C4 , explain the significant correlation observed genome wide between methylation and replication timing ( , ) ) [101] . Note that chromatin state compositional content in Monocd14ro1746 is quite the same as in K562 ( Supplementary Fig . S11 ) . In constrast , C3 and C4 in GM12878 have exchanged their GC and CpGo/e distributions ( Supplementary Fig . S9 ) . Interestingly , this phenomenon is paired with C3 becoming more late in GM12878 than C4 ( Supplementary Fig . S9 ) . This observation suggests that the genomic regions that replicate late in S phase are more likely specified by sequence features than by epigenetic features . However , the GC content cannot be the primary determinant of MRT for C1 and C2 states . Indeed the GC distributions in C1 and C2 are nearly the same ( Fig . 8A , Supplementary Fig . S9A and S11A ) whereas a great discrepancy is observed in the MRT distributions ( Fig . 6 , Supplementary Fig . S8 and MRT data non available ) . Once mapped on the genome ( Fig . 9A , B ) , the four prevalent chromatin states differ not so much in the genome coverage but mainly in their number and length distribution of domains or blocks of adjacent 100-kb-loci in the same chromatin state ( Table 2 and Fig . 9C ) . C1 and C2 chromatin blocks are more numerous but they are shorter with a mean length kb and 228 kb respectively . Their length pdfs do not reveal many domains larger than 1 Mb . C3 chromatin blocks are slightly less numerous and also mostly short , the larger mean length kb resulting from the existence of a few large C3 streches of several Mb length . The C4 block length pdf definitely differs from the previous ones by the presence of a fat tail . Not only the mean length kb is about three times the ones of C1 , C2 blocks , but most of the C4 domains exceed 1 Mb up to 5 Mb and more , hence they are less numerous ( Fig . 9C ) . This observation is quite consistent with the HP1-associated classical heterochromatin spreading mechanism and its possible association with the nuclear envelope [3] , [6] . When looking at the distribution of chromatin states along human chromosomes ( Fig . 9A , B ) , there is a clear evidence that C1 , C2 , C3 and C4 blocks are not distributed independently . In large regions with MRT ≲ 0 . 4 , short C1 and C2 blocks intersperse with each other , the C1s being the earliest ones ( e . g from 158 to 161 Mb in Fig . 9A ) . In a few 100 kb wide regions of MRT0 . 6 , C3 blocks are observed with a repressive effect ( e . g around 156 Mb in Fig . 9A where chromosome 1 contains a lot of olfactory receptor genes ) . C4 lies in very late regions MRT and form large uninterrupted blocks of several Mb size ( e . g from 185 to 190 Mb in Fig . 9A ) . This MRT dependent spatial organization of chromatin states prompted us to investigate neighborhood dependency between 100 kb loci . The obtained transition matrix ( Table 3 ) confirms that C4 loci have by far the highest probability ( 0 . 85 ) to have a C4 neighbor consistent with C4 blocks being much longer than the other chromatin state blocks ( Table 2 and Fig . 9C ) . It also quantifies the fact that C1 loci ( and in turn blocks ) have a much higher probability to have a neighbor that is a C2 locus ( block ) than a C3 or C4 locus ( block ) and vice-versa . This is consistent with the fact that C1 and C2 are likely to be replicated one after each other in early and mid S phase whereas C3 and C4 are replicated much later ( Fig . 6 ) . Consistently C4 loci ( blocks ) have a highest probability to have a neighbor that is a C3 locus ( block ) whereas C3 loci ( blocks ) have apparently no special preference . The spatial organization of chromatin blocks suggests that we can associate C1+C2 on one side and C3+C4 on the other side ( Supplementary Fig . S2B ) resulting in large-scale blocks of surprisingly very similar length distributions ( Fig . 9D ) with fat tails and respective means 779 kb and 808 kb . These mega-base long C1+C2 and C3+C4 chromatin blocks would on average be replicated rather early ( Fig . 9E ) and late ( Fig . 9F ) , respectively . Importantly , fixing the number of chromatin states to two in our PCA and cluster analysis does not result in the same dichotomic picture ( Supplementary Fig . S2A ) . Instead we discriminate the active chromatin state C1 from a composite silent state C2+C3+C4 ( Supplementary Fig . S2B ) Note that when using the so-computed transition matrix between chromatin states ( Table 3 ) to generate randomly synthetic chromosomes , we obtained very good predictions for the four chromatin state block mean lengths ( Table 2 ) . However the corresponding sample standard deviations so predicted are significantly smaller than the ones computed for the genuine human chromosomes which is an indication that the succession of chromatin states along human chromosomes is probably governed by a more global and elaborated underlying segmentation process . When concentrating our study on the 876 replication timing U-domains previously identified in K562 cells [80] , we revealed some remarkable organization of the four prevalent chromatin states ( Fig . 10A ) . The highly expressed gene rich euchromatin state C1 is found to be confined in a closed ( ) neighborhood of the “master” replication origins that border each individual U-domains ( Fig . 10A ) . As confirmed on the mean occupation profiles obtained for four U-domains size categories ( Fig . 10 E , F , G , H ) , this confinement is independent of the U-domains size and consistent with the previous observation [80] , [101] that U/N-domain borders are significantly enriched in DNase I hypersensitive sites and in insulator-binding proteins CTCF . C1 can thus be seen as specifying the early initiation zones that border U-domains and that were further shown [80] to delimit topological domains on genome-wide ( Hi-C ) chromatin state conformation data . The Pc repressed heterochromatin state C2 is mostly found at finite distance ( ∼200–300 kb ) from U-domain borders as clearly seen on the largest U-domains whose centers are drastically devoided of C2 loci ( Fig . 10B , H ) . In small U-domains ( ) , C2 occupies in majority their centers ( Fig . 10E , F ) that are replicated in mid-S phase . U-domain borders are also significantly depleted in unmarked and constitutive heterochromatin states C3 ( Fig . 10C ) and C4 ( Fig . 10D ) , respectively . C3 is already present in the center of small U-domains ( Fig . 10E , F ) and homogeneously occupies large U-domain centers ( Fig . 10G , H ) . C4 is significantly found in the center of U-domains that are larger than 1 Mb; C4 spreads and becomes predominant when increasing the size of U-domains beyond 1 . 8 Mb ( Fig . 10G , H ) . These results show that the replication “wave” starting from the early initiation zones at U-domain borders and propagating inside U-domains during S-phase with the progressive activation of secondary replication origins [79] , actually corresponds to a directional path through the four prevalent chromatin states C1 , C2 , C3 and ultimately C4 in the largest U-domains . This gradient of chromatin structure , from active openess at U-domain borders to closeness at U-domain centers via intermediate Pc repressed and unmarked heterochromatins is likely to be a key ingredient in the long-range chromatin control of the spatio-temporal replication program that underlies the megabase-sized replication fork polarity gradients observed in about 50% of the human genome [79] , [80] . In summary , this integrative analysis of epigenetic mark maps in the immature myeloid human cell line K562 has shown that the combinatorial complexity of these epigenetic data can be reduced to four prevalent chromatin states , one transcriptionally active open euchromatin state C1 and three distinct and silent heterochromatin states , namely a Pc repressed state C2 , a unmarked silent state C3 and a HP1-associated constitutive state C4 . By performing this statistical study at the ( low ) resolution 100 kb of available genome-wide MRT data , we have found that these chromatin states actually replicate at distinct periods of the S-phase , C1 replicates early , C2 is a mid-S phase phase state whereas C3 replicates later than C2 but before C4 that replicates very late , almost at the end of S-phase . In the Supplementary Data are reported , for comparison , the results of a similar integrative analysis of epigenomic data in the lymphoblastoid cell line GM12878 ( Supplementary Figs S6 , S7 , S8 and S9 ) and in the blood cell line Monocd14ro1746 ( Supplementary Figs S10 , S11 ) , which confirm that the classification of the human epigenome in four main chromatin states likely summarizes the data in different cell types . Interestingly , these four main chromatin states display remarkable similarities with that found in different cell types in Drosophila [21] and Arabidopsis [19] at the resolution ∼1 kb of gene expression data , suggesting the existence of simple principles of organization in metazoans as well as in plants [19]–[22] . When mapping these four chromatin states along the human chromosomes , our study reveals that the human genome can be segmented into megabase-sized domains of three different types with distinct spatio-temporal replication programs . In 50% of the human genome that are covered by the replication U-domains [80] , the U-shape of the replication timing profile indicates that the effective replication velocity ( which equals the inverse of the replication timing derivative [80] , [98] ) increases from U-domain borders to centers [79] as the signature of an increasing origin firing frequency during S-phase [132] . Our results ( Fig . 10 ) show that this acceleration of the replication wave is actually observed along a directional path through the four main chromatin states , the open euchromatin state C1 at U-domain borders successively followed by the heterochromatin states C2 , C3 and C4 at the U-domain centers . To which extent this chromatin gradient influences fork progression from the “master” early initiation zones at U-domain borders and secondary origins activation inside U-domains is a key issue of current modeling [79] , [133]–[135] of the spatio-temporal replication program in human and more generally in mammals . The complete analysis of the other half of the human genome that is complementary to U-domains is more in agreement with the traditional dichotomic picture proposed in early studies of the mouse [55]–[57] and human [59] , [65] , [75] genomes , where early and late replicating regions occur in separated compartments of open and close chromatin , respectively . About 25% of the human genome are covered by megabase sized GC-rich ( C1+C2 ) chromatin blocks that on average replicate early by multiple almost synchronous origins with equal proportion of forks coming from both directions ( Table 4 ) . This absence of well-positioned origins explains that the skew has not accumulated in these gene-rich regions that were shown to be devoided of skew N-domains [90]–[93] . The last 25% of the human genome corresponds to megabase sized GC-poor domains of interspersed ( C3+C4 ) heterochromatin states or of long C4 domains that on average replicate late by again multiple almost coordinated origins ( Table 4 ) . These gene-poor regions are also devoided of skew N-domains and can be seen as the late replicating counter-part of the gene-rich ( C1+C2 ) regions . Extending this study to different cell types including ES , somatic and cancer cells looks very promising . By performing our integrative analysis at low ( 100 kb ) and high ( 1 kb ) resolutions in parallel , we should be in position to investigate the global reorganization of replication domains during differentiation ( or disease ) in relation to coordinated changes in chromatin state and gene expression . For example , this multivariate approach should shed a new light on the so-called replication domain “consolidation” phenomenon [56] that corresponds to the disappearance ( EtoL transition ) or appearance ( LtoE transition ) of a U-domain border during differentiation [80] . The probable coordinated change in chromatin state at 100 kb resolution and the possible change at 1 kb resolution are likely to explain the possible change in gene expression . This opens new perspectives in the study of chromatin-mediated epigenetic regulation of transcription and replication in mammalian genomes in both health and disease .
Timing profiles for the immature myeloid cell line K562 and the lymphoblastoid cell line GM06990 were obtained from the authors [80] . The mean replication timing ( MRT ) is given for 27656 100 kb non-overlapping windows in hg18 coordinates . We also retrieved the coordinates of the 876 U-domains in K562 and 882 U-domains in GM06990 from the authors [80] . For all ChIP-seq data , we downloaded data in the Encode standard format “broadpeaks” ( http://genome . ucsc . edu/FAQ/FAQformat . html ) . Broadpeaks format is a table of significantly enriched genomic intervals . Most of the data correspond to the release 3 ( August 2012 ) of the Broad histone track . We downloaded the tables from: http://hgdownload . cse . ucsc . edu/goldenPath/hg19/encodeDCC/wgEncodeBroadHistone/ . The CBX3 and Sin3A data corresponds to the release 3 ( September 2012 ) of the HAIB TFBS track . Tables were downloaded from the UCSC from: http://hgdownload . cse . ucsc . edu/goldenPath/hg19/encodeDCC/wgEncodeHaibTfbs/ For the K562 cell line , we downloaded the broadpeak tables for the following antibodies: CTCF , H3K27ac , H3K27me3 , H3K36me3 , H3K4me3 , H3K9me3 , RNAP ll , H2AZ , H3K79me2 , H3K9me1 , H4K20me1 , CBX3 , Sin3A . For the GM12878 cell line , we downloaded: CTCF , H3K27ac , H3K27me3 , H3K36me3 , H3K4me3 , H3K9me3 . For the Monocd14ro1746 cell line we downloaded: CTCF , H2AZ , H3K27ac , H3K27me3 , H3K36me3 , H3K4me3 , H3K79me2 , H3K9ac , H3K9me3 . Genomic intervals were then mapped back to hg18 using LiftOver . For each ChIP-seq data , we computed a profile at the 100 kb resolution for the 27656 non-overlapping windows for which MRT is defined . The read density for one antibody in a window is the number of reads in this window that fall in significantly enriched intervals normalized by the window length . All statistical computations were performed using the R software ( http://www . r-project . org/ ) . In order to compute the Spearman correlation matrix , the epigenetic profiles at 100 kb resolution were transformed with the R function rank with option ties . method = max . Then we computed the Pearson correlation matrix on the transformed dataset . To reorder the matrix in Fig . 1 , we computed the Spearman correlation distance as: ( 1 ) where is the spearman correlation . Then , a dendrogram was computed using the R function hclust with option method = average and with as dissimilarity . Principal component analysis was performed on the rank transformed dataset using the function dudi . pca from the R package ade4 ( see http://pbil . univ-lyon1 . fr/ADE-4 and Ref . [107] ) with the option scale = TRUE ( i . e . each variable is centered and normalized before the PCA computation ) . The first three components were retained which accounts for of the dataset variance ( see Supplementary Fig . S1 ) , and clustering was performed in this 3D space . We used Clara algorithm [88] which is an optimization of k-means for large data set . We used the clara function implemented in the R package cluster . The options were set to: stand = FALSE , sampsize = 500 , samples = 20 , metric = euclidean . To assess the number of clusters , we used the pooled within-cluster sum of squares around the cluster mean . Suppose that the data set of size is divided in k clusters . Let d ( x , y ) be the euclidean distance between the points x and y . Let be the mean of the cluster , then the within-cluster sum of squares for this cluster is: ( 2 ) The pooled within-sum of squares for the k clusters is: ( 3 ) The pooled within-cluster sum of squares necessarily decreases with the number of clusters . A good choice for the number of clusters is the critical point where some clear crossover is observed from a fast decrease of at small k values to a weak decrease of at large k values . This means that , after this critical point , no much information is gained by adding a new cluster . In our analysis this crossover occurs for k = 4 clusters ( see Fig . 3B ) . We also used the Gap statistic [109] which is defined by : ( 4 ) is the expected value of for a sample of size drawn from a proper reference distribution . We choose , as a reference , a uniform distribution over the range of the observed data . A good choice for the number of clusters is a value of k so that is much smaller than the expected from a random distribution ( i . e . a high value of ) . Four clusters is also a reasonable choice according to the gap statistic index computed with R package clusterSim ( see Fig . 3C ) . Poorly clustered data points were removed from the set of chromatin states . The silhouette value [108] is a way to quantify how well a point is clustered . Definition 1 . Given a particular clustering , , of the data in k clusters , let i be a data point and the average distance of the data point i to the members of the cluster . Let i be a member of cluster and ( 5 ) The silhouette value of the data point i is defined as: ( 6 ) A silhouette value below 0 means that the data point is actually closer in average to the points from another cluster than to the one it has been assigned to . Points with a negative silhouette value are border line allocations . We decided to remove those points from the set of identified chromatin states . Hence chromatin states are groups ( clusters ) with homogeneous epigenetic features . 91% of all 100 kb non-overlapping windows of the human genome were assigned to one of the four chromatin states C1 , C2 , C3 or C4 . The number of transitions from i to j , , is the number of 100 kb windows of state i contiguous to a window of state j ( the sense or antisense orientation is not taken in account ) . Let be the number of windows in chromatin state i . The conditional probability of a transition from i to j given i is . As human gene coordinates , we used the UCSC Known Genes table . When several genes presenting the same orientation overlapped , they were merged into one gene whose coordinates corresponded to the union of all the overlapping gene coordinates , resulting in 23818 distinct genes . Expression data were retrieved from the Genome Browser of the University of California Santa Cruz ( UCSC ) . To construct our expression data set , we used RefSeq Genes track as human gene coordinates . Genes with alternative splicing were merged into one transcript by taking the union of exons . Hence the TSS was placed at the beginning of the first exon . We obtained a table of 23329 genes . We downloaded expression valuess from the release 2 of Caltech RNA-seq track ( ENCODE project at UCSC: http://hgdownload . cse . ucsc . edu/goldenPath/hg18/encodeDCC/wgEncodeCaltechRnaSeq/ ) . Expression for one transcript is given in reads per kilobase of exon model per million mapped reads ( RPKM ) [136] . RPKM is defined as: ( 7 ) where C is the number of mappable reads that fall into gene exons ( union of exons for genes with alternative splicing ) , N is the total number of mappable reads in the experiment , and L is the total length of the exons in base pairs . We associated 17872 genes with a valid RPKM value in K562 . CpG observed/expected ratio ( CpG o/e ) was computed as , where , and are the numbers of C , G and dinucleotides CG , respectively , counted along the sequence , L is the number of nonmasked nucleotides and l is the number of masked nucleotide gaps plus one , i . e . L-l is the number of dinucleotide sites . The CpG o/e was computed over the sequence after masking annotated CGIs . The GC content was computed on the native sequence . We detected contiguous windows of the same chromatin state ( C1 to C4 ) . We then kept the coordinates of the blocks of contiguous windows . To form chromatin state blocks of states ( 1+2 ) , we merely detected contiguous windows of state 1 or 2 . The same procedure was applied to define chromatin blocks of states ( 3+4 ) . For chromatin blocks ( 1+2 ) and ( 3+4 ) , we authorized the inclusion of isolated windows which don't belong to any chromatin state so to not disrupt very long blocks . Each gene name of our annotation dataset was associated to several GO terms from GO SLIM ( high level GO terms ) using the online mapper: http://go . princeton . edu/cgi-bin/GOTermMapper . Then for each chromatin state ( C1 to C4 ) , the number of occurrences of each GO term was determined by the number of promoters belonging to that state and associated to this GO term . The enrichment for each GO term in each cluster was tested using Fisher's exact test . We applied a procedure to control the false discovery rate ( FDR ) as described in [137] . The upper limit of the FDR was fixed to 20% . After detecting significant deviation from a random repartition of GO term occurrences , we used the odd ratio value to determine if the deviation was an enrichment ( ) or a depletion ( ) . | Previous studies revealed spatially coherent and biological-meaningful chromatin mark combinations in human cells . Here , we analyze thirteen epigenetic mark maps in the human cell line K562 at 100 kb resolution of MRT data . The complexity of epigenetic data is reduced to four chromatin states that display remarkable similarities with those reported in fly , worm and plants . These states have different MRT: ( C1 ) is transcriptionally active , early replicating , enriched in CTCF; ( C2 ) is Polycomb repressed , mid-S replicating; ( C3 ) lacks of marks and replicates late and ( C4 ) is a late-replicating gene-poor HP1 repressed heterochromatin state . When mapping these states inside the 876 replication U-domains of K562 , the replication fork polarity gradient observed in these U-domains comes along with a remarkable epigenetic organization from C1 at U-domain borders to C2 , C3 and ultimately C4 at centers . The remaining genome half displays early replicating , gene rich and high GC domains of intermingled C1 and C2 states segregating from late replicating , gene poor and low GC domains of concatenated C3 and/or C4 states . This constitutes the first evidence of epigenetic compartmentalization of the human genome into replication domains likely corresponding to autonomous units in the 3D chromatin architecture . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Human Genome Replication Proceeds through Four Chromatin States |
The Drosophila melanogaster gene archipelago ( ago ) encodes the F-box/WD-repeat protein substrate specificity factor for an SCF ( Skp/Cullin/F-box ) -type polyubiquitin ligase that inhibits tumor-like growth by targeting proteins for degradation by the proteasome . The Ago protein is expressed widely in the fly embryo and larva and promotes degradation of pro-proliferative proteins in mitotically active cells . However the requirement for Ago in post-mitotic developmental processes remains largely unexplored . Here we show that Ago is an antagonist of the physiologic response to low oxygen ( hypoxia ) . Reducing Ago activity in larval muscle cells elicits enhanced branching of nearby tracheal terminal cells in normoxia . This tracheogenic phenotype shows a genetic dependence on sima , which encodes the HIF-1α subunit of the hypoxia-inducible transcription factor dHIF and its target the FGF ligand branchless ( bnl ) , and is enhanced by depletion of the Drosophila Von Hippel Lindau ( dVHL ) factor , which is a subunit of an oxygen-dependent ubiquitin ligase that degrades Sima/HIF-1α protein in metazoan cells . Genetic reduction of ago results in constitutive expression of some hypoxia-inducible genes in normoxia , increases the sensitivity of others to mild hypoxic stimulus , and enhances the ability of adult flies to recover from hypoxic stupor . As a molecular correlate to these genetic data , we find that Ago physically associates with Sima and restricts Sima levels in vivo . Collectively , these findings identify Ago as a required element of a circuit that suppresses the tracheogenic activity of larval muscle cells by antagonizing the Sima-mediated transcriptional response to hypoxia .
Metabolically active tissues require an adequate supply of dioxygen ( O2 ) for metabolic production of ATP by aerobic glycolysis and as a necessary substrate in a variety of enzymatic reactions ( reviewed in [1] ) . Consequently , cells in metazoan organisms have evolved a conserved hypoxia-response mechanism that senses low O2 ( or hypoxia ) and modulates cellular metabolism and signaling in response to this environmental challenge . Activation of this adaptive mechanism results in changes in transcription that allow organisms to adapt to O2 conditions that might otherwise be incompatible with normal development and homeostasis ( reviewed in [2] ) . In most metazoans , these changes include elevated expression of factors involved in oxygen-independent ATP production , increased expression of oxygen-carrying hemoglobin-like molecules and increased branching of O2-carrying tubular organs , the net effect of which is to reduce overall O2 demand and increase O2 delivery . Molecular mechanisms through which changes in O2 concentration alter metabolism and drive increased tubular branching are conserved through the metazoan tree to include invertebrates like the fruit fly Drosophila melanogaster ( reviewed in [3] ) . A key element of this mechanism is the hypoxia-inducible factor-1 [4]–[7] ( HIF-1 or Drosophila HIF [dHIF] in flies ) , which is a heterodimeric transcription factor composed of an oxygen-regulated HIF–1α subunit and a constitutively expressed HIF–1β subunit . In Drosophila these subunits are respectively encoded by the similar ( sima ) [8] , [9] and tango ( tgo ) [10]–[12] genes . The HIF-1/dHIF heterodimer is required for cellular adaptation to hypoxic conditions [4]–[7] and is regulated mainly at the level of HIF–1α stability [reviewed in 13] ) . In normoxic conditions , HIF–1α is hydroxylated at conserved proline residues by the 2-oxoglutarate/Fe ( II ) -dependent prolyl-4-hydroxylase family member HIF prolyl hydroxylase ( HPH ) [14] , [15] . Prolyl-hydroxylation of HIF–1α facilitates binding with the von Hippel Lindau ( VHL ) E3-ubiquitin ligase subunit , and subsequent polyubiquitination and proteasome-dependent degradation of HIF-1α [16]–[20] . Drosophila Sima is controlled by a well-conserved version of this pathway involving the HPH homolog fatiga ( fga ) , and the Drosophila VHL homolog , dVHL [14] , [21]–[24] . Because the HPH enzymatic activity is dependent upon the availability of oxygen [14] , [15] , the HPH/VHL pathway effectively functions as a sensor of cellular oxygen levels , allowing HIF–1α/Sima stabilization only in hypoxic conditions and preventing HIF activity in normoxic cells [reviewed in 2] . Mutations in the VHL gene stabilize HIF-1α and are associated with a dominantly inherited hereditary cancer syndrome in humans that predisposes to a variety of malignant and benign tumors of the eye , brain , spinal cord , kidney , pancreas , and adrenal glands [25] . Excess HIF-1α can promote several important aspects of cancer biology , including the metabolic switch to anaerobic glycolysis characteristic of tumor cells [i . e . the Warburg effect; 26] , neoangiogenesis , and increased tumor metastasis [reviewed in 13] , [27] , [28] . The invertebrate response to hypoxia mirrors key features of the mammalian hypoxic response [3] , [29] , [30] . Hypoxia stabilizes Sima and induces expression of genes that include homologs of mammalian HIF targets , such as lactate dehydrogenase ( LDH ) [31] . Hypoxic treatment of flies also produces physiological changes reminiscent of the mammalian hypoxic response [32] , including altered metabolism and reduced oxygen consumption [33]–[36] . Adult Drosophila respond to hypoxia by entering into state of stupor characterized by low or undetectable neurological activity that allows them to tolerate extended periods of low oxygen [34] , and recovery from this state is dependent upon genes necessary for survival in low-oxygen conditions [31]–[33] , [35] . Hypoxia also induces a neoangiogenesis-like process in Drosophila involving increased branching of the tracheal system , an open network of interconnected , epithelial tubes that duct gases in and out of the animal [reviewed in 37] . Drosophila larvae reared in chronic hypoxia show increased branching of cells at the tip of each tracheal branch termed ‘terminal tip’ cells , whereas those raised in chronic hyperoxia show a reciprocal decrease in the extent of terminal branch elaboration [22] , [38] . This increased larval tracheal branching in low O2 involves the FGF receptor homolog breathless ( btl ) [39] and the FGF ligand branchless ( bnl ) [40]: hypoxic exposure results in a sima-dependent increase in expression of btl in tracheal cells and bnl in peripheral oxygen-deficient tissues [22] , [38] . Bnl then acts on tracheal terminal tip cells , which express Btl [41] , [42] , to induce fine tubular extensions that project toward Bnl-expressing cells . These terminal branches serve as the primary site of gas exchange between the tracheal system and internal tissues . When the oxygen demand is met , Bnl and Btl expression decreases , thereby limiting hypoxia-induced tracheal growth . This oxygen responsiveness allows for growth of tracheal terminal branches specifically to localized areas of hypoxia in order to shape the mature tracheal architecture and to increase oxygen-delivery capacity in hypoxic conditions . In addition to the oxygen-dependent HPH/VHL pathway , mammalian HIF-1 is regulated by VHL-independent mechanisms that are incompletely understood [43] , [44] . Recent studies have linked HIF–1α turnover to phosphorylation by the GSK3ß kinase and subsequent binding of the ubiquitin ligase subunit Fbw7 [45] , [46] , which is a sequence and functional ortholog of the Drosophila Archipelago ( Ago ) protein . Intriguingly Drosophila Ago binds and stimulates turnover of the Trachealess protein ( Trh ) , which is a Sima/HIF-1α sequence homolog , in embryonic tracheal cells [47] . Genetic data show ago and dVHL also coregulate oxygen-sensitivity in the developing embryonic tracheal arbor [48] . In light of these connections , we have tested the requirement for ago in oxygen-sensitive stages of larval tracheal development and find evidence that ago is an antagonist of dHIF during the larval stage . Genetic manipulations that reduce ago function within post-mitotic larval muscle cells lead to a sima-dependent increase in the branching of nearby terminal cells . This phenotype is not suppressed by a trh allele that suppresses branch defects in ago mutant embryonic tracheal cells [47] , but rather correlates with elevated expression of the Sima-induced gene bnl expression in larval muscle cells and genetic dependence on bnl . At an organismal level , reducing ago activity results in constitutive expression of some dHIF target genes in normoxia , increases the sensitivity of others to mild hypoxic stimulus , and allows adult flies to recover more rapidly from hypoxic stupor than normal flies . Significantly , non-cell autonomous effects of ago and dVHL alleles on terminal branching are synergistic , suggesting that the Ago and dVHL proteins co-regulate dHIF . Consistent with this , Ago protein can be found in a complex with Sima in larval extracts and loss of Ago elevates Sima levels in peripheral tissues . Collectively these findings define an important role for Ago as a required antagonist of the Sima-dependent hypoxic response during the larval stage of Drosophila development .
Heterozygosity for a null allele of ago sensitizes the Drosophila embryonic tracheal system to mild hypoxia [48] . To determine whether ago is also involved in hypoxia responsiveness in the subsequent larval stage , it was necessary to generate an allele of ago that allowed development beyond the late embryonic lethality associated with ago null alleles [49] . This was achieved by transposase-mediated imprecise excision of EP ( 3 ) 1135 , a P-element located 16 base pairs ( bp ) upstream of the ago genomic locus ( Bloomington Drosophila Stock Center [BDSC] ) that behaves genetically as a weak ago hypomorph . Excision of EP ( 3 ) 1135 produced a 603 bp deletion removing the first exon of the ago-RC transcript ( Figure 1A–1B ) that was designated agoΔ3–7 . The effect of agoΔ3–7 on patterns of ago transcription was determined by quantitative real-time PCR ( qRT-PCR ) . Of the three predicted ago transcripts ( ago-RA , ago-RB , and ago-RC ) only RA and RC are detected in whole larvae ( Figure 1C ) . Consistent with the location of the deletion in the agoΔ3–7 allele , the ago-RC transcript is specifically absent in agoΔ3–7 mutant larvae while expression of the ago-RA transcript is unaffected . Notably , the ago-RA and RC transcripts display inverse expression patterns: ago-RA is approximately 3-fold more abundant than ago-RC in imaginal discs and larval brain and ventral nerve cord , but ago-RC is 3-fold more abundant than ago-RA in filleted larval body wall preparations ( Figure 1D ) . The agoΔ3–7 allele is thus a tissue- and transcript-specific allele that primarily reduces ago expression in peripheral tissues such as body wall muscle . Approximately 49% of agoΔ3–7 homozygotes or trans-heterozygotes in combination with the null alleles ago1 and ago3 die as pupae ( Table 1 ) and the remainder die as late 2nd and 3rd instar larvae ( data not shown ) . Those that live to late 3rd instar show tracheal phenotypes ( Figure 2 and Table 2 ) . The most prevalent of these phenotypes is an approximate doubling of the number of cytoplasmic branches elaborated from multiple subtypes of terminal cells , including those found along the lateral trunk that serve to oxygenate the ventrolateral body wall muscles: LH cell terminal branching increases from 20 . 4±0 . 64 branches ( n = 33 ) in control larvae to 39 . 5±1 . 59 branches ( n = 34 ) in agoΔ3–7/1 larvae ( p = 2 . 6×10−16 ) ( Figure 2A–2C and Table 2 ) , and LG cell terminal branching increases from 19 . 6±0 . 54 branches ( n = 33 ) in control larvae to 36 . 8±1 . 94 branches ( n = 31 ) in agoΔ3–7/1 larvae ( p = 9 . 5×10−13 ) ( Figure 2C ) . Notably , the magnitude of these increases in terminal cell branch number is similar to that seen in larvae grown in hypoxic conditions [22] , [38] . Loss of ago function also causes additional tracheal branch phenotypes in approximately 25% of larvae , including the appearance of terminal branch tangles ( Figure 2D ) and the development of ‘ringlet’-shaped ganglionic branches ( Figure 2F ) , which resemble phenotypes seen in hypoxic larvae or those in which Sima is activated by genetic disruption of the Fga/dVHL regulatory pathway [22] . Given the transcript- and tissue-specific nature of the agoΔ3–7 allele , these tracheal phenotypes support the hypothesis that ago has a non-autonomous role in in restricting terminal branching . Although the agoΔ3–7/1 larval phenotypes are reminiscent of hypoxia-induced tracheal growth , they do not exclude the possibility that an earlier developmental requirement for ago ( e . g . in the embryo ) affects later branching events in the larva . To test the temporal requirement for ago in regulating tracheal terminal branching patterns , a dominant negative ago transgene ( UAS-agoΔF ) [47] , [50] was combined with the hs-Gal4 driver to produce animals in which ago activity could be antagonized at later developmental stages by application of a heat-shock . Whereas control and hs>agoΔF larvae show similar LH cell branch number prior to transgene induction ( 22 . 2±0 . 89 branches [n = 27] vs 21 . 7±0 . 69 branches [n = 24] ) , administration of a transient heat-shock to hs>agoΔF larvae is sufficient to drive an increase in terminal branching throughout the tracheal system ( effects on LG and LH cells quantified in Figure 3A and 3B ) . LH cell branching is increased 24 hrs post heat-shock in hs>agoΔF larvae ( 40 . 2±1 . 48 branches [n = 24]; ( p = 3 . 0×10−14 relative to no heat-shock ) but remains unchanged in control larvae ( 22 . 6±0 . 67 branches [n = 24] ) . Thus animals that complete embryonic and early larval development with wt ago activity can be induced to undergo excess branching by transient expression of an ago dominant-negative allele . Excess terminal branch phenotypes in hs>agoΔF and agoΔ3–7 animals may reflect a requirement for ago in either tracheal or non-tracheal cell types . To directly test whether ago activity is required in non-tracheal tissue to restrict branching , the agoΔF transgene was driven with the 5053A-Gal4 driver ( 5053A>agoΔF ) , which is expressed specifically in ventrolateral body wall muscle 12 ( VLM12 ) and has been used to study non-cell autonomous tracheogenic activity of the Btl/Bnl pathway [38] . The VLM12 muscle expresses endogenous , nuclear Ago protein ( Figure 4C–4D ) and is normally tracheated by the LF and LH cells ( Figure 4A ) . The 5053A>agoΔF genotype approximately doubles the number of LF and LH tracheal branches that terminate on VLM12 ( Figure 4B ) relative to either the adjacent muscle ( VLM13 ) or to control larvae expressing a nuclear-localized GFP ( nlsGFP ) ( 5 . 11±0 . 16 branches [n = 54] in control vs 9 . 54±0 . 29 branches [n = 50] in agoΔF , p = 4 . 67×10−24 ) ( Table 3 ) . This degree of excess branching produced by muscle-specific expression of the agoΔF transgene is similar to that produced by organism-wide depletion of the Ago-RC isoform with the agoΔ3–7 genomic allele . These combined genetic data provide evidence that Ago is required within larval body wall muscle cells to restrict the post-embryonic branching of nearby tracheal terminal cells . ago mutations lead to tissue-specific activation of factors normally degraded by the SCF-Ago ubiquitin ligase , including the proliferative proteins CycE and dMyc in larval imaginal discs [49] , [50] and the transcription factor Trachealess in tracheal cells [47] . Although the expression patterns of these proteins in body wall muscle are not well defined , we wished to test whether ectopic expression of CycE , dMyc , Trh , or the SCF-Fbw7 target Notch [reviewed in 51] was even capable of conferring a non-cell autonomous tracheogenic activity to VLM12 . To this end , each of these factors was individually overexpressed using the 5053A-Gal4 driver ( Table 3 ) . Muscle-specific expression of trh or Notch failed to stimulate excess terminal branch growth . The inability of trh to affect tracheal recruitment to VLM12 contrasts with its ability to phenocopy ago mutant phenotypes in the embryonic trachea [47] and further suggests that the ago larval tracheal role is from separable from its embryonic role . Muscle-specific expression of dMyc also had no effect on the degree of terminal cell branching , despite a 28 . 5% increase in the 2-dimensional size of the VLM12 muscle ( Table 4 ) . Notably , increased tracheation of VLM12 driven by agoΔF occurs without an increase in the size of the VLM12 muscle , which is consistent with no role for post-mitotic growth in this phenotype ( Table 4 ) . 5053A>cycE does increase terminal branch number , although to a lesser degree than agoΔF . However , CycE protein levels are not obviously affected by expression of agoΔF ( Figure S1 ) , suggesting that deregulated CycE is an unlikely cause of the non-autonomous effect of ago alleles on terminal cell branching . The similarity of ago mutant terminal branching phenotypes to those induced by hypoxia suggests that ago may antagonize the dHIF pathway . To test the genetic relationship between ago and sima in larval tracheal branching , the sima07607 loss-of-function allele [23] was introduced into the 5053A>agoΔF and agoΔ3–7/1 genetic backgrounds . Heterozygosity for sima ( i . e . sima07607/+ ) dominantly suppressed the agoΔF VLM12 phenotype ( Table 3 ) by decreasing terminal branch number from 9 . 54±0 . 29 ( n = 50 ) to 6 . 34±0 . 27 branches ( n = 53 , p = 4 . 36×10−12 ) , and also suppressed the excess and overlapping terminal branching seen in agoΔ3–7/1 larvae ( Table 2 and Figure 5A–5B; white arrow in 5A indicates a ringlet-shaped ganglionic branch ) from 39 . 5±1 . 59 ( n = 34 ) to 29 . 0±1 . 48 branches per LH cell ( n = 34 , p = 7 . 45×10−6 ) . In addition , the sima07607 allele dominantly delayed the lethal phase of both agoΔ3–7 homozygotes and agoΔ3–7/1 or agoΔ3–7/3 trans-heterozygotes ( Table 1 ) . Reciprocally , ectopic expression of sima in the VLM12 muscle ( 5053A>sima ) increased tracheal recruitment in normoxic conditions ( Table 3 ) . Muscle cells are thus distinct from ectodermal cells , which do not recruit branching following overexpression of sima [22] . To more directly assess dHIF activity in ago mutant animals , the transcription of the Drosophila LDH gene ( dLDH ) was measured in the body wall muscle of agoΔ3–7 and control larvae . LDH is a well-validated HIF target in vertebrates and invertebrates , and HIF-responsive elements from the LDH promoter have been used as the basis of HIF activity reporters in many different systems including Drosophila [e . g . 24] . This analysis showed a 27 . 3-fold increase in dLDH transcription in ago mutant larval body wall muscle preparations but no equivalent upregulation in steady-state levels of the sima mRNA ( Figure 5C ) . Sima-driven expression of the FGF ligand bnl is a key element of the hypoxic response among non-tracheal cells [22] , [38] , [52] . The bnlP1 loss-of-function allele dominantly suppressed both the 5053A>agoΔF VLM12 phenotype ( Table 3 ) , from 9 . 54±0 . 29 ( n = 50 ) to 6 . 54±0 . 28 branches ( n = 54 , p = 5 . 99×10−11 ) and the agoΔ3–7/1 excess branching phenotype ( Table 2 ) , from 39 . 5±1 . 59 ( n = 34 ) to 28 . 4±1 . 80 branches per LH cell ( n = 29 , p = 1 . 75×10−5 ) . In parallel , qRT-PCR detected an ∼50% upregulation of bnl transcription in body wall muscle of agoΔ3–7 larvae relative to control muscle ( Figure 5C ) . Previous studies using a genomic duplication of the bnl locus have demonstrated that a similar 50% increase in bnl gene-dosage is sufficient to elicit excess tracheal terminal cell branching [38] . Thus reduced ago function in body wall muscle is associated with ectopic expression of the dHIF target dLDH , increased levels of the bnl mRNA , and a genetic dependence on sima and bnl . The data above suggests that ago alleles might exhibit functional interactions with components of the Fga/dVHL pathway , which controls Sima stability and activity in vivo [21]–[23] , [53] , [54] . A previously characterized dVHL RNAi knockdown transgene ( dVHLi ) [48] ) was used with the 5053A-Gal4 driver to reduce dVHL expression in VLM12 . Consistent with the role of dVHL upstream of sima , the 5053A>dVHLi genotype showed an increase in terminal branching relative to a non-specific RNAi control ( Figure 6A , and Table 3 ) ( 5 . 28±0 . 20 branches [n = 40] in Adf1i control vs 7 . 48±0 . 21 branches [n = 89] in dVHLi , p = 1 . 27×10−9 , Figure 6 ) . The dVHLi and agoΔF transgenes were then co-expressed with 5053A-Gal4 to determine their ability to enhance VLM12 tracheogenic activity ( Figure 6C–6D ) . The 5053A>agoΔF , VHLi compound genotype shows a synergistic increase in the number of branches that terminate on VLM12 ( Table 3 ) , but also leads to a phenotype not seen in either individual genotype: whereas expression of agoΔF or dVHLi individually increase terminal branching of LF and LH onto VLM12 , the agoΔF+dVHLi combination also recruits ectopic branches from the LG lateral terminal cell ( as seen in the two different focal planes of a single agoΔF+dVHLi-expressing VLM12 muscle; Figure 6C–6D ) which normally bypasses VLM12 . This ectopic LG recruitment phenotype occurs in approximately 10% of agoΔF+dVHLi VLM12 muscles and is also observed upon 5053A-Gal4 driven expression of bnl [38] or sima ( data not shown ) . Thus dVHL and ago are individually required to restrict the ability of muscle cells to recruit new branch growth , and combined reduction of ago and dVHL activity leads to increased tracheogenic signals emanating from body wall muscle . To further define the relationship between ago and dVHL in terminal branching , transgenes expressing each factor were tested for rescue of VLM12-branching phenotypes produced by reducing the function of the other ( Table 3 ) . Expression of wild type dVHL led to a 66% suppression of the agoΔF branching phenotype ( p = 6 . 55×10−12 ) ; reciprocally , over-expression of wild type ago showed a 54% suppression of the dVHLi branching phenotype ( p = 2 . 73×10−4 ) . Thus , each gene can to some degree ameliorate non-autonomous branching phenotypes produced by loss of the other in the VLM12 segment . In view of the genetic and molecular links between ago , sima , dLDH , and dVHL , the organism-wide transcriptional response to hypoxia was examined in ago mutant animals . Drosophila respond to varying degrees of hypoxia by driving transcription of distinct sets of target genes at differing oxygen concentrations , including those involved in metabolic adaptation and survival in low oxygen [31] , [32] . A subset of hypoxia-inducible genes was selected for this analysis based on their differential transcription in hypoxic adult Drosophila [31] and predicted links to known mechanisms of the hypoxic response . These included dLDH , which plays a role in the metabolic switch to high flux glycolysis [reviewed in 55] , [56] , lysyl oxidase ( lox ) , a HIF target in mammalian cells that plays a role in hypoxia-induced changes in cell adhesion [57] , [58] and vascular remodeling [59] , and dHIG1 ( CG11825 ) , the Drosophila homolog of Hypoxia Induced Gene-1 ( HIG1 ) , which is induced by HIF and promotes cell survival [60] . qRT-PCR analysis was carried out for each of these genes under conditions of decreasing environmental oxygen ( 21% , 5% , 0 . 5% ) in whole control larvae or whole agoΔ3–7 larvae ( Figure 7A–7B ) . We find that each of these genes is differentially induced in hypoxia in a manner consistent with findings in adult Drosophila [31] and can be ectopically induced by the agoΔ3–7 allele . dLDH is minimally transcribed in normoxic control larvae , and with progressively higher transcription as the oxygen level falls ( 1 . 6 and 2 . 7-fold increases in 5% and 0 . 5% O2 respectively , Figure 7A ) , confirming that dLDH transcription increases with increasing dHIF activity . In agoΔ3–7 homozygous animals , dLDH expression is increased 8 . 1-fold in whole normoxic larvae ( this lower fold induction in the whole larva relative to the ∼27-fold enrichment seen in dissected body wall muscle in Figure 5C is presumably a reflection of the tissue-specific nature of the agoΔ3–7 allele ) , and is increased approximately 14-fold in agoΔ3–7/1 larvae relative to control larvae at both 5% and 0 . 5% O2 ( Figure 7B , top panel ) . Thus ago restricts dLDH expression activity across a broad range of oxygen concentrations . The lox gene is normally only up-regulated in whole control larvae by strong hypoxia ( 4 . 4-fold induction at 0 . 5% O2; Figure 7B ) . The agoΔ3–7 allele leads to a 2 . 2-fold increase in lox transcription in normoxia , and lox transcription reaches near maximal levels at 5% O2; the 3 . 4-fold induction seen in ago mutants in 5% O2 is not significantly different from that seen in control larvae at 0 . 5% O2 ( Figure 7B , middle panel ) . This pattern suggests that the lox promoter is induced by levels of dHIF activity achieved in moderate hypoxic conditions , and that this threshold is more easily reached in ago mutants . The dHIG1 gene displays a more exaggerated version of the lox response pattern: dHIG1 mRNA levels are only induced strongly ( 19 . 9-fold ) in whole control larvae by 0 . 5% O2 ( Figure 7A ) ; the agoΔ3–7 allele is not sufficient to drive ectopic dHIG1 transcription in normoxic conditions but it is sufficient to sensitize the dHIG1 promoter to reduced O2 levels such that maximal dHIG1 expression is now achieved at a ten-fold higher O2 concentration than normal ( Figure 7B , bottom panel ) . In addition to dLDH , lox , and dHIG1 , three other genes also induced by hypoxia , hairy , amy-p and thor genes [31] , are also moderately up-regulated in normoxic agoΔ3–7 mutant larvae ( Table S1 ) . Reducing ago activity is thus sufficient to alter the threshold required to drive expression of multiple hypoxia-inducible genes . Adult Drosophila respond to prolonged periods of oxygen deprivation by entering into a state of hypoxic stupor characterized by inactivity and reduced oxygen consumption [34] . Many mutations have been identified that slow this hypoxic recovery [32] , [33] , [35] , but few mutations have been described that enhance it . Under our standard laboratory conditions , control adult flies enter stupor after approximately fifteen to twenty minutes in a 0 . 5% O2 environment and remain unconscious until re-oxygenation . We assayed control +/+ adults , agoΔ3–7/+ adults , and adults trans-heterozygous for the agoΔ3–7 allele and the ago hypomorphic allele EP ( 3 ) 1135 ( BDSC ) for recovery time following acute hypoxia ( 1 hour at 0 . 5% O2 ) ( Figure 7C ) . agoΔ3–7/EP ( 3 ) 1135 flies display no obvious developmental phenotypes and enter into hypoxic stupor at the same rate as control flies ( data not shown ) ; however , they recover significantly faster than either control +/+ or agoΔ3–7/+ adults . Linear regression analysis indicates that the time for 50% recovery is reduced from 4 . 5±0 . 75 minutes in control +/+ flies , to 1 . 4±0 . 16 minutes in agoΔ3–7/EP ( 3 ) 1135 flies ( p = 0 . 0015 ) . The agoΔ3–7/EP ( 3 ) 1135 population also reaches 100% recovery after 10 minutes of re-oxygenation , whereas neither the control +/+ or agoΔ3–7/+ populations achieved 100% recovery by the end of the 15 minute measurement period ( data not shown ) . Thus , the genetic evidence of a role for ago as a regulator of dHIF-regulated branching in the larval tracheal arbor is paralleled at the organismal level by an enhanced transcriptional sensitivity to hypoxia and an increased ability of flies to recover from a transient hypoxic challenge . To test the molecular relationship between Sima and Ago , Sima levels were assessed in two ways: by immunoflourescent staining of VLM12 muscles expressing the UAS-agoΔF transgene and by Western blotting of lysates of agoΔ3–7 larvae ( Figure 8 ) . Fluorescence microscopy confirms that a previously described anti-Sima antibody [24] detects high levels of transgenically expressed Sima in the VLM12 nuclei of 5053A>sima muscles , and that endogenous Sima is not readily detectable by this method of analysis in the nuclei of adjacent non-transgenic muscles ( Figure 8A ) . Following expression of the agoΔF dominant-negative transgene ( 5053A>agoΔF ) , a fraction of VLM12 nuclei accumulate anti-Sima reactive epitopes ( see arrows , Figure 8B ) . This same anti-Sima antibody detects elevated levels of an ∼110 kD molecular weight band in agoΔ3–7 filleted 3rd instar pelts relative to wt control pelts ( Figure 8C ) . This ∼110 kD band is absent in lysates of sima07607 larvae ( Figure 8D , lane 1 vs . 2 ) , and is specifically enriched in precipitates of an anti-Ago polyclonal antibody from lysates of hypoxic larvae ( Figure 8D , lane 5 ) . Collectively , these molecular data indicate that Ago can associate with Sima in larval lysates , and that Ago limits Sima levels in developing tissues .
The selective stabilization of the Sima/HIF–1α transcription factor in hypoxia plays a key role in the response of metazoan organisms to low oxygen concentrations by its ability to induce a program of hypoxia-specific gene expression [reviewed in 2] . Evidence suggests that in Drosophila , Sima plays a dual role in the post-mitotic growth of tracheal terminal branch cells toward hypoxic peripheral tissues by acting within both the ‘signaling’ hypoxic peripheral cells and in the ‘responsive’ terminal tips cells [reviewed in 37] . Our data implicate the Ago-SCF ubiquitin ligase as a required regulator of Sima during hypoxia sensing in peripheral cells , but do not rule out an additional role for Ago within tracheal terminal tip cells which contributes to their ectopic branching in ago mutant larvae ( see below ) . Phenotypes produced by muscle-specific expression of an ago dominant-negative allele , or by a genomic allele that specifically affects ago expression in peripheral tissues , are phenocopied by overexpression of Sima ( this study ) or the FGF homolog Bnl [38] . These non-cell autonomous effects of ago alleles on terminal branching are accompanied by a strong induction in peripheral tissues of the dHIF target dLDH , and can be dominantly suppressed by an allele of sima . ago alleles induce expression of a set of dHIF-inducible hypoxia-response genes in normoxia that includes dLDH , and this is paralleled at the organismal level by an enhanced ability of ago mutant flies to the recover from hypoxic stupor . ago alleles are thus among the first genetic alterations shown to enhance the recovery of adult Drosophila from hypoxic exposure . Within larval muscle , ago appears to inhibit sima in parallel to dVHL , which targets the Sima/HIF–1α protein for constitutive degradation in normoxia [reviewed in 3] . Consistent with this , we find evidence that Ago can associate with Sima and limits its levels in vivo . Collectively these data significantly expand the known role of Ago in organism development by demonstrating that it is required in an apparently novel pathway that collaborates with dVHL to inhibit Sima-regulated hypoxic gene expression in peripheral tissues . Though the work presented here focuses on the ‘tracheo-attractant’ effects of reducing ago expression in body wall muscle , this may be just one manifestation of roles Ago plays in controlling hypoxia-regulated gene expression . Indeed , reducing ago function has a quantitatively stronger effect on terminal branching than a genomic duplication of the bnl locus [38] , suggesting either that ago also act within tracheal cells to limit branching [as in 47] , [48] or that a larger set of dHIF target genes contribute to the effect . Consistent with this latter hypothesis , normoxic ago mutant larvae display ectopic induction of hypoxia-responsive metabolic genes such as dLDH , lox , hairy , amy-p and thor . Based on this profile , it appears that ago mutant larvae reared in normoxia elevate expression of bnl but also engage a metabolic switch to high-flux glycolysis that is characteristic of hypoxic cells [32] , [33] , [35] , [36] , [61] . Future studies will be required to assess the full effect of these transcriptional changes on the behavior of terminal tracheal cells and the tissues into which they project . In wild type animals , the transcriptional response of cells to hypoxia is graded such that different target genes are induced across a range of environmental O2 concentrations [31] . In ago mutants , this differential induction is largely abolished such that expression of genes such as lox and dHIG1 is virtually indistinguishable at 5% and 0 . 5% O2 . Thus , ago appears to be required both for inhibition of hypoxia-inducible genes in normoxia and for the graded expression of hypoxia-inducible genes under variable levels of oxygen deprivation . We hypothesize that this graded sensitivity is normally a product of the interaction between the Ago and Fga/dVHL regulatory mechanisms . The HPH/VHL pathway has been demonstrated to act in a graded manner , such that it degrades HIF–1α efficiently in normoxia , but is progressively less efficient as the oxygen concentration drops [62] . This leads to a gradient of HIF activity that is presumably required for the differential induction of target genes . We hypothesize that ago acts in parallel to dVHL to dampen Sima/HIF-1 activity across a range O2 concentrations , and that Ago may function as a dHIF regulatory mechanism at very low O2 concentrations in which the HPH/dVHL pathway is hypothesized to be inactive [62] . Thus , the absence of Ago allows a mild hypoxic stimulus ( ∼e . g . 5% O2 ) to be translated into levels of dHIF-dependent gene expression that would normally only result from much stronger hypoxic exposure . The data presented here support this prediction , with the end result that the transcriptional response profile of hypoxia-response genes in ago mutant larvae is shifted toward induction by more mild stimuli . The molecular mechanism ( s ) underlying the genetic relationship between ago and sima in tracheal branching appears to involve a physical association of their encoded proteins that modulates Sima levels . Given that Ago is a ubiquitin ligase specificity factor , these data are consistent with a model in which Ago supports Sima polyubiquitination and turnover . Recent studies have identified the human Ago ortholog Fbw7 as a HIF–1α interacting factor and have proposed that Fbw7 promotes HIF-1α turnover following GSK3ß phosphorylation in cultured cells [45] , [46] . Phenotypic predictions made by this molecular model appear to be confirmed by the ago terminal cell branching phenotypes documented here . Intriguingly , RNAi depletion of GSK3ß/shaggy or a proteasome subunit in VLM12 also elevates the number of tracheal branches that terminate on this muscle ( Table S2 ) . However Fbw7 is implicated in the proteolytic destruction of two transcription factors , the Notch intracellular domain ( NICD ) [reviewed in 51] and sterol-regulatory enhancer binding protein ( SREBP ) [63] , that indirectly modulate HIF-dependent hypoxic gene expression in eukaryotic cells [64] , [65] . Ago could thus theoretically influence hypoxic gene expression via these paths as well . Future biochemical studies will be required to clarify the full range of Ago molecular targets that contribute to its role in hypoxic gene expression . The well-studied anti-proliferative role of ago is conserved in its human ortholog Fbw7 , which is mutationally inactivated in a wide spectrum of primary human cancers [reviewed in 51] . Some cancer cells engage a program of gene expression that supports a switch to high-flux glycolysis ( a phenomenon termed ‘Warburg effect’ [26] ) and are more resistant to transient hypoxia than normal cells [reviewed in 1] . Both of these properties can now , to some degree , also be associated with ago loss in Drosophila . In view of the functional conservation between SCF-Ago and SCF-Fbw7 in degradation of shared oncogenic targets [49] , [50] and the proposed role of Ago/Fbw7 in Sima HIF-1α turnover [this study and 45] , [46] , our data raise the interesting possibility that sensitization to mild hypoxia may be a feature of Fbw7 mutations in vertebrates as well . If so , then tumor suppressive properties of Fbw7 may derive in part from its established anti-proliferative role and in part due to modulation of HIF-regulated angiogenic and metabolic pathways .
The FRT80B and w1118 strains were used as wild type controls . The ago1 and ago3 alleles have been previously described [49] . The agoΔ3–7 allele was identified as an imprecise excision of the agoEP ( 3 ) 1135 transposon . Alleles used in this study: bnlP1 , sima07607 , agoEP ( 3 ) 1135 ( all from the Bloomington Drosophila Stock Center ) and 1-eve-1 [66] . The following transgenes were also used: UAS-agoΔF and UAS-ago [47]; UAS-CycE , UAS-dMyc , UAS-N , UAS-nlsGFP ( all from the Bloomington Drosophila Stock Center ) , UAS-trh [67] , UAS-sima [24] , UAS-dVHL [21] , UAS-Adf1RNAi and UAS-sggRNAi ( Vienna Drosophila RNAi Center ) , UAS-dVHLRNAi [48] , hs-Gal4 , and 5053A-Gal4 ( from the Bloomington Drosophila Stock Center ) . Statistical comparisons were made using Student's t-Test ( Microsoft Excel ) with the indicated significance levels . Hypoxia treatments were performed in a sealed Modular incubator chamber ( Billups-Rothenberg Inc . , Del Mar , CA ) with separate gas intake and exhaust openings . Internal O2 concentration was measured with an electronic O2 sensor ( OX-01 , RKI Instruments , Inc . , Union City , CA ) . To assay recovery from hypoxia , 5–7 day old adult flies were put into plain glass tubes in groups of 9–15 . The flies were then placed into the hypoxia chamber at 0 . 5% O2 for one hour and then removed to normoxia . Following hypoxic treatment , >99% of the flies ( 178 of 179 ) had fallen into hypoxic stupor . Recovery time was defined as the time required for each individual fly to resume walking following re-oxygenation . Total RNA was isolated from dissected third instar larval body wall muscles . cDNA was reverse-transcribed using random hexamer primers ( Invitrogen ) with Superscript II Reverse Transcriptase ( Invitrogen ) . dVHL and ß-tubulin transcripts were then amplified with gene-specific primers . For quantification of mRNA levels , total RNA was isolated from whole third instar larvae or dissected larval tissues and reverse transcribed as described above . Levels of Arp87c , ago-RA , -RB and -RC , dLDH , sima , bnl , lox , hairy , dHIG1 , thor and amy-p were then assayed with gene-specific primers using the SYBR green method of quantitative real-time PCR on a Roche LightCycler 480 machine . Transcript abundance was normalized to levels of Arp87c as in [31] . To image the larval tracheal system , third instar larvae were dissected in cold PBS and fixed in 4% paraformaldehyde . Air-filled tracheal branches were imaged using bright-field microscopy . and assembled using Photomerge ( Adobe Photoshop CS ) . Third instar larvae were dissected in cold PBS , fixed in 4% paraformaldehyde and incubated with guinea pig anti-CycE ( 1∶500 ) or rabbit anti-Sima ( 1∶1000 ) . Secondary antibodies ( anti-guinea pig conjugated to Cy3 or anti-rabbit conjugated Cy5 ) were used as recommended ( Jackson ImmunoResearch ) . To assess Sima protein levels in third instar larvae , larval pelt extracts were prepared in sample buffer and resolved on 7 . 5% SDS-PAGE prior to Western blotting with rabbit anti-Sima ( 1∶1000 ) [24] and developed with anti-rabbit HRP ( 1∶1000; Jackson ImmunoResearch ) . Whole larval extracts were immunoprecipitated with guinea pig anti-Ago polyclonal sera ( 1∶1000 ) [47] prior to immunoblotting with anti-Sima antibody . | Cells in multicellular animals must adapt to changing environmental conditions in order to ensure survival of the larger organism . One key challenge they face is fluctuation in the availability of dissolved oxygen . As cells get low on oxygen , they respond by turning on a program of gene expression that helps them survive . The key to this program is a protein , called HIF-1α in humans and Similar ( or Sima ) in the fruit fly Drosophila melanogaster , that is kept inactive in normoxia but is activated in hypoxia . The mechanisms responsible for this switch are not completely understood . In this study , we present genetic and molecular evidence that a component of the protein degradation machinery called Archipelago is required to keep Sima inactive in developing muscle cells and that genetically removing Archipelago makes these cells “think” they are hypoxic . This finding and the data that support it provide new insight into genetic circuits that cells use to control their response to changing oxygen levels and suggest that defects in oxygen homeostasis may contribute to cancerous disease states associated with loss of the human equivalent of Archipelago called Fbw7 . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"growth",
"control",
"cellular",
"stress",
"responses",
"cancer",
"genetics",
"developmental",
"biology",
"morphogenesis",
"gene",
"expression",
"biology",
"molecular",
"biology",
"signal",
"transduction",
"cell",
"biology",
"cell",
"migration",
"genetics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2013 | The Archipelago Ubiquitin Ligase Subunit Acts in Target Tissue to Restrict Tracheal Terminal Cell Branching and Hypoxic-Induced Gene Expression |
The guanine nucleotide exchange factor Vav1 is essential for transducing T cell antigen receptor signals and therefore plays an important role in T cell development and activation . Our previous genetic studies identified a locus on rat chromosome 9 that controls the susceptibility to neuroinflammation and contains a non-synonymous polymorphism in the major candidate gene Vav1 . To formally demonstrate the causal implication of this polymorphism , we generated a knock-in mouse bearing this polymorphism ( Vav1R63W ) . Using this model , we show that Vav1R63W mice display reduced susceptibility to experimental autoimmune encephalomyelitis ( EAE ) induced by MOG35-55 peptide immunization . This is associated with a lower production of effector cytokines ( IFN-γ , IL-17 and GM-CSF ) by autoreactive CD4 T cells . Despite increased proportion of Foxp3+ regulatory T cells in Vav1R63W mice , we show that this lowered cytokine production is intrinsic to effector CD4 T cells and that Treg depletion has no impact on EAE development . Finally , we provide a mechanism for the above phenotype by showing that the Vav1R63W variant has normal enzymatic activity but reduced adaptor functions . Together , these data highlight the importance of Vav1 adaptor functions in the production of inflammatory cytokines by effector T cells and in the susceptibility to neuroinflammation .
The guanine nucleotide exchange factor ( GEF ) Vav1 is essential for transducing T cell antigen receptor ( TCR ) signals and therefore plays a critical role in the development and activation of T cells [1–5] . Following TCR engagement , Vav1 becomes rapidly tyrosine phosphorylated by kinases of the Src and/or Syk family . This phosphorylation relieves Vav1catalytic Dbl homology ( DH ) domain and causes Vav1 to promote GDP-GTP exchange on the Rho , Rac1 and Cdc42 small GTPases [6 , 7] . In addition to this GEF activity , the CH , SH2 and SH3 domains of Vav1 allow its incorporation into the LAT-Grb2-Gads-PLCγ1-SLP76 micro-clusters that form at the immunological synapse . Binding of Vav1 appears to stabilize this molecular complex thereby controlling PLCγ and nuclear factor of activated T cells ( NFAT ) activation [1 , 8 , 9] . In Vav1-deficient mice , T cell development is partially blocked at the pre-TCR checkpoint in the thymus and both positive and negative selections of T cells are strongly impaired [3 , 5] . Furthermore , TCR-induced activation and proliferation is greatly diminished in Vav1-deficient T cells , due to reduced TCR-induced signaling that impacts on Ca2+ flux , on the activation of Erk , protein kinase D1 ( PKD1 ) and serine-threonine kinase Akt , and on the activity of transcription factors such as NFAT and nuclear factor κB ( NF-κB ) [1 , 2 , 10] . A recent study has shown that many critical events involved in T cell activation are mediated by either the GEF or the scaffolding activities of Vav1 [11] . The GEF activity of Vav1 is necessary for T cell development and for the optimal activation of T cells , including signal transduction to Rac1 , Akt , and integrins . In contrast , Vav1 GEF activity is not required for TCR-induced Ca2+ flux , activation of Erk and PKD1 , cell polarization and development of regulatory T cells ( Treg ) . Lewis ( LEW ) and Brown-Norway ( BN ) rats behave in opposite ways concerning their susceptibility to autoimmunity , allergy and infectious diseases [12 , 13] . Genetic dissection using these rat strains has identified a locus of 117 kb on chromosome 9 that controls natural Treg development [14] . Fine mapping of this locus revealed a non-synonymous SNP in the Vav1 gene leading to the substitution of an arginine residue by a tryptophan at position 63 ( R63W ) in BN rats . Interestingly , this 117 Kb interval is fully included in the Eae4 locus of 1 cM that controls the susceptibility to central nervous system ( CNS ) inflammation [15] . Although this study suggested that Vav1 could be involved , one important limitation was the possibility that other genetic variants contained in the 117 Kb fragment besides the Vav1R63W polymorphism could be responsible for these phenotypes . Here , we sought to unequivocally test the involvement of the Vav1R63W polymorphism in the susceptibility to CNS inflammation and to determine its mechanisms of action . To this aim , we generated a knock-in mouse model in which the arginine at position 63 was replaced by a tryptophan residue . Using this model , we show that Vav1R63W mice display reduced susceptibility to experimental autoimmune encephalomyelitis associated with a lower production of effector cytokines by autoreactive CD4 T cells that is intrinsic to effector CD4 T cells . Finally , we provide a mechanism for the above phenotype by showing that the Vav1R63W variant has normal enzymatic activity but reduced adaptor functions . Together , these data highlight the importance of Vav1 adaptor functions in the production of inflammatory cytokines by CD4 T cells and in the susceptibility to neuroinflammation .
The development of T cells in the thymus proceeds through a series of stages that are controlled by signals triggered by the pre-TCR and TCR complexes . We analyzed the impact of the Vav1R63W polymorphism on these stages by using Vav1R63W mice ( S1 Fig ) . Vav1R63W mice displayed a consistent lower cellularity in the thymus compared to that of wild-type littermates , with significantly fewer CD4+CD8+ double positive ( DP ) , CD4+CD8- ( CD4SP ) or CD4-CD8+ ( CD8SP ) single positive cells ( Fig 1A ) . In contrast , Vav1R63W mice had similar absolute numbers of CD4-CD8- double negative ( DN ) thymocyte subsets ( Fig 1B ) , suggesting normal pre-TCR signaling . The expression of CD5 , a negative regulator that correlates positively with TCR signal intensity , was significantly lower on Vav1R63W DP thymocytes ( Fig 1C ) indicating an impaired TCR signaling . To better investigate the positive selection of thymocytes towards the CD4 lineage , we crossed Vav1R63W to OT-II transgenic mice , which express an ovalbumin-specific TCR restricted to MHC class II I-Ab . We observed that the generation of thymocytes expressing the ovalbumin-specific TCR was impaired in Vav1R63W mice ( Fig 1D ) . Similar results were obtained using female HY-TCR transgenic mice whose Vβ6+ TCR recognizes a male-specific antigen presented by MHC class II I-Ab ( Fig 1E ) . Next , we investigated the effect of the Vav1R63W mutation on negative selection . As previously reported [16] , male HY-TCR transgenic mice displayed a dramatic reduction of DP cells in the thymus due to early stage negative selection . In contrast , male HY-TCR transgenic mice in a Vav1R63W background had twice as many thymocytes , resulting from increased DP cells ( Fig 1F ) . These data indicate that negative selection is also impaired in Vav1R63W KI mice . Collectively , our results reveal that Vav1R63W has no major effect on the pre-TCR checkpoint but rather causes a defect in TCR-driven positive and negative selections of DP thymocytes . Although Vav1R63W mice exhibit a defect in thymic development , there were no significant differences in the proportion and absolute numbers of CD4 and CD8 T cells in lymph nodes ( Fig 2A ) and spleen ( S2A Fig ) . Yet , flow cytometry analysis of CD4 T cells from Vav1R63W mice revealed a slightly higher frequency of CD44highCD62low T cells , suggesting an increased in T cells with effector/memory phenotype . We hypothesized that this could be attributed to homeostatic proliferation—also known as lymphopenia-induced proliferation [17]—resulting from decreased thymic output . To determine whether the Vav1R63W mutation affected effector CD4 T cell functions , we investigated the proliferation and cytokine production of sorted naïve CD4+CD62Lhigh T cells after stimulation with anti-CD3 and anti-CD28 mAbs in vitro . Proliferation analysis revealed only mild differences ( S2B Fig ) . In contrast , CD4 T cells from Vav1R63W mice produced less IFN-γ and TNFα , but more IL-4 ( Fig 2B ) . We also examined the development of Treg cells in the lymphoid organs and observed a significant increase in the proportion of CD4+Foxp3+ Treg cells in the thymus , spleen and lymph nodes of Vav1R63W mice ( Figs 2C and S2C ) . We next investigated the impact of the Vav1R63W mutation on the in vitro suppressive activity of CD4+Foxp3+ Treg cells . Freshly isolated CD4+CD62L+CD25- naïve T cells were stained with cell trace violet and cultured with soluble anti-CD3 mAb and antigen-presenting cells for 3 days in the presence or absence of sorted CD4+CD62L+CD25bright Treg cells . More than 95% of sorted CD4+CD62L+CD25bright cells expressed Foxp3 . Effector CD4 T cells stimulated in the absence of Treg cells proliferated readily and the immunosuppressive potential of Treg cells from Vav1R63W mice was preserved at various Treg/Teff cell ratios ( Fig 2D ) . Moreover , Treg cells from WT and Vav1R63W mice expressed similar levels of Foxp3 , CTLA-4 , GITR and PD1 ( S2D Fig ) . Thus , our results show that the Vav1R63W mutation impacts on the cytokine production of effector T cells but not on the suppressive function of Treg cells in vitro . We previously reported that the Vav1R63W polymorphism is fully included in the 1 cM locus controlling susceptibility to CNS inflammation in rats [15] . To formally demonstrate the implication of the Vav1R63W polymorphism in this phenotype , WT and Vav1R63W mice were immunized with 50 or 100 μg of MOG35-55 peptide . Although the incidence of the disease was similar between the two groups , Vav1R63W mice developed a less severe disease , with delayed onset and quicker recovery resulting in reduced clinical scores ( Fig 3A ) . The analysis of CNS infiltration at day 15 and 30 after MOG35-55 immunization revealed no significant differences in the numbers of CD4 T cells or CD4+Foxp3+ T cells infiltrating the brain ( Figs 3B and S3A ) or the spinal cord ( Figs 3C and S3B ) , thereby excluding a defect in CD4 T cell migration or increased Treg numbers as the basis of reduced disease severity . However , the CNS-infiltrating CD4 T cells from both the brain ( Fig 3B ) and spinal cord ( Fig 3C ) of MOG35-55-immunized Vav1R63W mice produced significantly less IFN-γ , IL-17 and GM-CSF . Similar results were obtained in draining lymph nodes ( Fig 4A ) . Using tetramer staining , we showed similar frequencies of MOG35-55-specific CD4 T cells in the brain ( Fig 3D ) and draining LNs ( Fig 3E ) of WT and Vav1R63W mice , although a moderate decrease was observed in the spinal cord of Vav1R63W mice ( Fig 3D ) . This suggests that the reduced cytokine expression is not the consequence of impaired development or expansion of MOG-specific CD4 T cells . We next examined whether the defect in cytokine production observed in Vav1R63W mice is intrinsic to effector CD4 T cells or is the consequence of either increased Treg frequency or modified function of other immune cells such as APCs . For this purpose , we generated mixed bone marrow chimeras by transferring a 1:1 mixture of bone marrow from WT mice bearing the congenic CD45 . 1 marker and Vav1R63W mice bearing both CD45 . 1 and CD45 . 2 congenic markers into irradiated lymphopenic CD45 . 2 recipient mice . These chimeras revealed that the cytokine profile of Vav1R63W CD4 T cells upon MOG35-55 immunization was not influenced by the presence of hematopoietic cells from WT mice ( Fig 4B ) . In addition , we showed that reduced EAE course in Vav1R63W mice was not changed by the depletion of Treg by one single injection at day 17 of the anti-CD25 PC61 mAb which depletes Tregs ( Fig 4C ) . Altogether , our results highlight the key role of Vav1 in the pathophysiology of EAE and suggest that the Vav1R63W polymorphism protects against the development of CNS inflammation by reducing the production of encephalitogenic cytokines by autoreactive CD4 T cells . We next investigated which Vav1-dependent TCR signaling pathways were affected by the Vav1R63W mutation . We first showed that the Vav1R63W variant was highly phosphorylated , together with a 75% reduction of its expression at the protein level ( Fig 5A ) . Vav1R63W had , however , no impact on proximal TCR signaling , as revealed by normal phosphorylation levels of ZAP70 , LAT and Lck ( Fig 5B ) . In contrast , this variant strongly impaired Vav1-dependent distal TCR signaling , as evidenced by a significant reduction in the phosphorylation of Erk , Akt and p38 ( Fig 5C ) . Further , this was associated with reduced calcium flux after TCR engagement ( Fig 5D ) . In contrast , the TCR induced activation of Rac1 was normal in Vav1R63W mice , suggesting normal GEF activity ( Fig 5E ) . Therefore , the biological effects of the Vav1R63W variant are likely to be mediated by the reduction in Vav1 protein expression and the consequent decrease in Vav1 adaptor functions .
In the present study , we analyzed the impact of the recently identified Vav1R63W variant on the development and functions of T cells , as well as on the susceptibility to CNS neuroinflammation . In contrast to Vav1 deficiency [1 , 3 , 5] , we found that the Vav1R63W variant had only mild effects on thymic development of T cells and on T cell homeostasis in the periphery . In addition , the Vav1R63W KI mice were less susceptible to CNS inflammation , resulting from a reduced production of inflammatory cytokines ( IFN-γ , IL-17 and GM-CSF ) by autoreactive CD4 T cells . Despite an increased proportion of Foxp3 Treg cells in Vav1R63W mice , the reduction in cytokine production was intrinsic to effector CD4 T cells and depletion of Treg cells had no impact on EAE development . Finally , we showed that Vav1R63W had normal GEF activity but reduced adaptor functions . Together , the analysis of this natural Vav1 variant formally established for the first time that Vav1 adaptor functions are essential for both T cell functions and susceptibility to autoimmune neuroinflammation . The interaction between the TCR and MHC-peptide complexes leads to the initiation of TCR signaling and represents a key step for the orchestration of the adaptive immune response . Indeed , the intracellular signaling pathways triggered upon TCR engagement finely control the thymic ontogeny of T cells , the mature T cell differentiation , expansion and activation [18 , 19 , 20] . In the thymus , engagement of the pre-TCR leads to the differentiation of the most immature DN thymocytes towards the DP stage , through a selection process called β-selection . Next , DP thymocytes expressing a mature TCRαβ and displaying low avidity for self-peptide-MHC complexes undergo a process of positive selection into either MHC class II-restricted CD4+CD8- or MHC class I-restricted CD4-CD8+ SP cells . Conversely , DP thymocytes that have a TCR with high avidity for self-peptide-MHC are eliminated by negative selection . The outcome of these different selection events is critically dependent on TCR signaling . Studies of Vav1-deficient mice have shown that the development of T cells is partially blocked at the pre-TCR checkpoint in the thymus , leading to a strongly block of both positive and negative selections [1 , 3 , 5] . In contrast , our results reveal that Vav1R63W has no major effect on the pre-TCR checkpoint , but rather causes a defect in TCR-driven positive and negative selections of DP thymocytes . In agreement , the critical downstream mediators of Vav1 signaling in response to TCR stimulation such as Ca2+ flux and activation of Erk , p38 and Akt were reduced in Vav1R63W CD4 T cells . In contrast , the phosphorylation of upstream signaling molecules such as LAT , LCK and ZAP70 was not affected . Since knock-in mice carrying a GEF-deficient Vav1 mutant revealed that the GEF activity of Vav1 is dispensable for Ca2+ flux and Erk activation [11] , our results suggest that Vav1R63W exhibits a defect in its adaptor functions . In contrast , the TCR induced activation of Rac1 was normal in Vav1R63W mice . Thus , this study highlights the essential roles of Vav1 adaptor functions in TCR induced positive and negative selections and its minor role in pre-TCR β-selection . We previously reported that the difference in susceptibility to EAE in rats was genetically controlled by a locus of 1 cM in chromosome 9 that contains the Vav1R63W polymorphism [14 , 15 ] . The present study using Vav1R63W KI mice provides the definitive demonstration that this mutation per se leads to a reduced severity of EAE . The analysis of CNS infiltration revealed no major differences in numbers of infiltrating CD4 T cells , thereby excluding the hypothesis that the reduced neuroinflammation observed in Vav1R63W KI mice may be the result of a defect in autoreactive CD4 T cell numbers or migration . Rather , the CD4 T cells from MOG35-55 immunized Vav1R63W KI mice produced significantly less inflammatory cytokines such as IFN-γ , IL-17 and GM-CSF . This defect likely originates in the periphery since similar results were obtained in draining lymph nodes . The analysis of MOG35-55-specific CD4 T cells using tetramer staining suggests that the reduced cytokine expression is not the consequence of impaired development or expansion of MOG-specific CD4 T cells . By using mixed bone marrow chimeras , we showed that the defect of cytokine production was intrinsic to effector Vav1R63W KI CD4 T cells and was not the consequence of either increased Treg cell frequency or modified function of other immune cells such as APCs . Consequently , depletion of Treg in Vav1R63W KI mice has no impact on EAE severity . This defect in cytokine expression is particularly relevant when considering that several studies have established that IFN-γ , IL-17 and GM-CSF are the main effector cytokines in the pathophysiology of both EAE and multiple sclerosis [21–23] . In addition , our findings are fully in line with our previous study using cohorts of patients with multiple sclerosis , in which we demonstrated a strong association between Vav1 expression , susceptibility to multiple sclerosis and production of inflammatory cytokines by CD4 T cells [15] . Our results , however , contrast with data using Vav1-deficient CD4 T cells or CD4 T cells harboring a mutated Vav1 with defective GEF activity , which rather showed increased production of IFN-γ [11 , 24 ] . The signaling pathway that depends on Vav1 adaptor function , therefore , plays an important role in Th1/Th17 differentiation and could be targeted for immunomodulation of immune mediated diseases . The differentiation of naïve CD4 T cells into functionally polarized T helper cell subsets depends notably on the strength of TCR-dependent signaling pathways upon antigen recognition [19] . In general , weak TCR signaling leads to transient Erk activation and favors Th2/Treg differentiation , whereas stronger TCR signaling leads to sustained Erk activation and favors Th1/Th17 differentiation [19 , 25 , 26] . Our findings are in line with these data , since we showed that Vav1R63W leads to a reduction of TCR signaling as revealed by a decrease of calcium flux and of Erk , Akt and p38 activities , which was associated with a decreased production of IFN-γ , IL-17 and GM-CSF by CD4 T cells . Of note , it was shown that Erk inhibitors could attenuate EAE by suppressing autoantigen-specific Th17 and Th1 responses [27] . Moreover , the genetic ablation of Erk2 impedes Th1 differentiation , while enhancing the development of induced Treg [28] . Concerning the p38 MAPK pathway , it has been shown that a single copy disruption of the p38 gene or a p38 inhibitor markedly reduce the pathogenesis of EAE by decreasing IL-17 production [29 , 30] . In contrast , the role of Akt-dependent pathways and calcium flux in CD4 T cell differentiation remains contradictory [31–35] The capacity to produce knockout mice has dramatically accelerated our knowledge on the immunological consequences of the complete loss of critical components of the TCR signaling pathways . However , in humans , the largest source of genetic variation is rather composed of single-nucleotide substitutions , for which it is far more difficult to predict their physiological consequences . Importantly , these can often affect biological pathways in unpredictable ways , as revealed recently by the use of mouse models with hypomorphic variants for SLP76 , LAT and Zap70 [36–39] . Our study reveals that the Vav1R63W model described herein is instrumental to expand our understanding of the immunological consequences of genetic variations of Vav1 expression and function . This model highlights the importance of Vav1 adaptor functions in the differentiation of CD4 T cells into Th1/Th17 subsets and suggest that genetic or acquired alterations in Vav1 signaling could play a major role in susceptibility to the many immune-mediated diseases , including autoimmune diseases where Th1/Th17 play a preponderant role .
Vav1 protein is evolutionarily conserved from nematodes to mammals and the analysis of rat and mouse Vav1 sequences revealed 98% homology . Comparative genomics studies indicated that the arginine found at position 63 in LEW rats is highly conserved among species while the tryptophan at this position is peculiar to BN rats . A genomic fragment containing exon 1 of the Vav1 gene was isolated from a BAC clone of C57BL/6 origin . The CGG codon found in exon 1 of the Vav1 gene and coding for the arginine residue present at position 63 of Vav1 was converted into a TGG codon coding for a tryptophane . A loxP-tACE-CRE-PGK-gb2-neor-loxP cassette ( NEO; [38] ) was introduced in the intron separating exons 1 and 2 of the Vav1 gene ( S1A Fig ) . After electroporation of JM8 . F6 C57BL/6N ES cells [40] and selection with G418 , colonies were screened for homologous recombination by Southern blot . The 3’ single-copy probe corresponded to a 582 bp genomic fragment located in intron 1 ( denoted 3’ probe in S1A Fig ) . When tested on BglI digested DNA , it hybridizes to a 16 . 4 kb wild-type fragment or to a 10 . 2 kb recombinant fragment . The occurrence of an appropriate homologous recombination event at the 5’ side was screened by PCR using the following oligonucleotides: 5’-AAACCTAGTGGGCGCTCTCCA-3’ and 5’-TGACGAGTTCTTCTGAGCGG-3’ . This pair of primer amplifies a 3791 bp fragment . Finally , a neomycin-specific probe was used to ensure that adventitious non-homologous recombination events had not occurred in the selected ES clones . Mutant ES cells were injected into FVB blastocysts . Germline transmission led to the self-excision of the NEO selection cassette in male germinal cells . Screening of mice for the presence of the Vav1R63W mutation was performed by PCR using the following oligonucleotides 5’-TGTAGGGGGCATCTGTCTGTCTG-3’ and 5’-AAATACCCTGGAGACTGCAGCAG-3’ . This pair of primers amplifies a 203 bp band in the case of the wild-type allele and a 269 bp band in the case of the Vav1R63W allele ( S1C Fig ) . Mice homozygous for Vav1R63W were fertile , indicating that the Vav1R63W mutation did not affect the embryonic development or viability of the KI mice . Mice harboring the Vav1R63W mutation ( international strain designation C57BL/6-Vav1tm2Mal ) were kept on a C57BL/6 background . OVA-specific OT-II-TCR transgenic mice [41] and HY-TCR transgenic mice , whose CD4 T cells are specific for HY peptide presented by IAb [16] were kindly provided by Dr . Sylvie Guerder ( Centre de Physiopathologie Toulouse-Purpan ) and were backcrossed with Vav1R63W mice . All mice were housed under specific pathogen-free conditions at the INSERM animal facility ( Zootechnie UMS-006; accreditation number A-31 55508 ) , which is accredited by the French Ministry of Agriculture to perform experiments on live mice . All experimental protocols were approved by the local ethics committee and are in compliance with the French and European regulations on care and protection of the Laboratory Animals ( EC Directive 2010/63 ) . The mAbs used for flow cytometry were as follows: RM4-5 ( anti-mouse CD4 ) , 53–6 . 7 ( anti-mouse CD8α ) , IM7 ( anti-mouse CD44 ) , PC61 ( anti-mouse CD25 ) , MEL-14 ( anti-mouse CD62L ) , H57-597 ( anti-mouse TCR αβ ) , FJK-165 ( anti-mouse Foxp3 ) , 53–7 . 3 ( anti-mouse CD5 ) , A20 ( anti-mouse CD45 . 1 ) , 104 ( anti mouse CD45 . 2 ) , anti-mouse IL-17A , anti-mouse GM-CSF , anti-mouse IFN-γ . The fluorescent conjugated antibodies were purchased from e-Biosciences , BD Biosciences and Biolegend . Antibodies used for ELISA were: 11B11 ( anti-IL-4 ) , AN18 ( anti-IFN-γ ) , purified anti-mouse IL-17A , purified anti-mouse GM-CSF , Biotin anti-mouse IFN-γ ( XMG1 . 2 ) , Biotin anti-mouse IL-17A , Biotin anti-mouse GM-CSF . These antibodies were purchased from BD Biosciences . The BVD6-24G2 ( anti-mouse IL-4 Biotin ) is from e-Biosciences . The MOG35–55 ( MEVGWYRSPFSRVVHLYRNGK ) peptide was purchased from Polypeptide Laboratories ( San Diego , CA ) with a purity grade >95% . 8–12 week-old mice were immunized subcutaneously at the base of the tail with 50 μg or 100 μg of MOG35-55 peptide emulsified in CFA ( BD Difco , Franklin Lakes NJ ) containing 500 μg of Mycobacterium tuberculosis ( Strain H37 , Difco ) . For active EAE , mice were injected intravenously with 200 ng of pertussis toxin ( List Biological Laboratories , Campbell , CA ) at days 0 and 2 post-immunization . Clinical scores were recorded daily as follow: 0 , no signs of disease; 1 , loss of tone in the tail; 2 , hind limb paresis; 3 , hind limb paralysis; 4 , tetraplegia; 5 , moribund . Intraperitoneal injection with anti-CD25 mAb ( PC61 ) ( 500 μg/ml ) antibody was performed at day 17 after MOG35-55 peptide immunization . Mice were anesthetized with Ketamine and perfused with cold PBS . Brain and spinal cord were collected separately , homogenized and digested with collagenase D ( 2 . 5 mg/ml , Roche Diagnostics ) , Dnase I ( 10 μg/ml ) and TLCK ( 1 μg/ml , Roche , Basel , Switzerland ) for 30 min at 37°C . Cells were then washed , suspended in 37% Percoll , and layered on 70% Percoll . After a 20-minute centrifugation at 2000 rpm , the mononuclear cells were collected from the interface , washed and resuspended in culture medium . Isolated cells were counted using a hematometer and then stained in order to analyze the presence of different cell populations by flow cytometry . 3x105 CNS infiltrated cells were stimulated O/N with different concentrations of MOG35-55 ( 0 , 10 and 100 μg ) to analyze the cytokine expression by CD4 T cells using intracellular staining ( ICs ) . Similarly LN cells and splenocytes were stimulated with different concentrations of MOG35-55 ( 0 , 10 and 100 μg ) for 72 hours to investigate the cytokine expression using intracytoplasmic staining and ELISA . MOG38-49-I-Ab and CLIP-I-Ab tetramers were obtained from the NIH tetramer core facility ( Emory University , Atlanta , USA ) . Cells from the brain , spinal cord and LNs were incubated with tetramers for 2h at room temperature at a concentration of 0 . 03 mg/ml and then stained for surface markers for flow cytometry analysis . CD45 . 2 Vav1R63W recipient mice were irradiated ( 9 . 5 Gy ) the day before i . v . injection of a 1:1 mixture of bone marrow cells . Cells were harvested after flushing cells from tibias and femurs and 20x106 cells from WT ( CD45 . 1 ) and Vav1R63W ( CD45 . 1xCD45 . 2 ) were injected per mouse . The control groups received BM from either WT or Vav1R63W mice . 8 weeks later , the chimeras were immunized with MOG35-55 and the cytokine profile of MOG-specific donor cells was analyzed by intracytoplasmic staining on day 15 after immunization . To purify naïve CD4+CD62L+ T cells , CD4 T cells were negatively selected using Dynal cocktail antibodies supplemented with PC61 ( an anti-CD25 mAb ) to eliminate Treg cells . Cells were then positively selected with anti-CD62L beads on MS columns ( Miltenyi , Auburn CA ) according to the manufacturer’s instructions . For functional test , naïve CD4 T cells were stimulated with plate-bound anti-CD3 ( 3 μg/ml ) and soluble anti-CD28 ( 1 μg/ml ) . Cytokine production was analyzed by ELISA 48h hours after culture . For Treg suppression assays , CD4 T cells were stained with antibodies directed against TCR , CD4 , CD62L and CD25 . CD4+CD62L+CD25- ( effector T cells ) and CD4+CD62L+CD25bright ( regulatory T cells ) were sorted using FACSAria SORP ( BD Biosciences ) . The purity of these populations was higher than 95% . Effector and regulatory T cells were co-cultured at different ratio ( Treg:Teff , 1:1 , 1:2 , 1:4 , 1:8 , 1:16 ) in the presence of WT irradiated splenic antigen presenting cells and soluble anti-CD3 ( 1 μg/ml ) . Effector cells were stained with cell trace violet ( life technologies ) in order to track their proliferation . Proliferation was analyzed 3 days after culture by flow cytometry . Enzyme immunoassays were used to measure cytokines in culture supernatants . 96 well plates were coated for 2 h at 37°C with anti-IFN-γ , anti-IL-17 or anti-GM-CSF in carbonate buffer 0 . 05 M pH 9 . 6 . Culture supernatants or standards were incubated 2 h at 37°C . The plates were then incubated for 1h30 min with a secondary biotinylated antibody specific for each cytokine , followed by 20 min incubation with streptavidin-phosphatase alkaline at 37°C . Finally , plates were revealed by phosphatase alkaline substrate and absorbance was measured at 450/540 nm . For intracellular cytokine staining , cells were stimulated with different concentrations of MOG35-55 ( 0 , 10 and 100 μg ) and treated with monensin ( Golgiplug 1 μg/ml , BD Biosciences ) at 37°C , in a humidified 5% CO2 atmosphere for 4 h . After staining of surface markers ( TCR , CD4 and CD44 ) , cells were fixed and permeabilized with Cytofix/Cytoperm and Perm/Wash buffer ( e-Biosciences ) according to the manufacturer’s instructions . Cells were then incubated with antibodies recognizing cytokines ( IL-17 , IFN-γ , GM-CSF ) or isotype controls for 20 min and washed twice with Perm/Wash buffer before analysis . The production of IL-4 , TNFα , IL-10 and IFN-γ by naive CD4 T cells was assayed using Cytometric Bead Array cytokine kit ( BD Biosciences ) . Purified CD4 T cells were stimulated at 37°C in non-supplemented RPMI 1640 using preformed complexes of biotinylated anti-CD3 ( clone 145-2C11 , Biolegend , 30 μg/ml /107 cells ) , anti-CD4 ( GK1 . 5 , Biolegend , 30 μg/107 cells ) and streptavidin ( 15 μg/107 cells ) . Stimulation was stopped by the addition of twice-concentrated lysis buffer ( 100 mM Tris , pH 7 . 5 , 270 mM NaCl , 1 mM EDTA , 20% glycerol and 0 . 2% n-dodecyl-β-maltoside ) supplemented with protease and phosphatase inhibitors . After 10 min of incubation on ice , cell lysates were centrifuged at 20 , 000 x g for 15 min at 4°C . The protein concentration was first assessed using bradford assay and normalized . Proteins were then denatured in Laemmli buffer and analyzed by SDS-PAGE followed by Western blotting on PVDF membranes ( Immobilon ) . ECL Prime ( Amersham ) was used as revelation substrate . Signal intensity quantification was performed using ImageJ software ( 1 . 47v ) . Two loadings were performed simultaneously for each time point , one for phosphoproteins and the other for total proteins . A GAPDH staining was also performed on each blot as loading control . The blots were then scanned , bands of interest were quantified and background line was subtracted . To determine specific phosphorylation , the signal from phosphorylated bands was divided by the appropriate loading control and all values were expressed as a fold increased of bands intensity . Signals below detection were set at 0 . The antibodies used for biochemical studies were , anti-Vav1 ( C-14 ) and anti-P-Vav1 Tyr174 From Santa Cruz Biotechnology , anti-LAT ( 1D-1 ) from Thermo Scientific , anti-P-LAT ( pY226 ) from BD Pharmingen , anti-Akt , anti-P-Akt Ser473 ( D9E ) , anti-P-Erk1/2 Thr202/Tyr204 ( D13 . 14 . 4E ) , anti-Erk ( 3A7 ) , anti ZAP-70 ( D1C10E ) , anti-P-ZAP70 ( Tyr319 ) , HRP-linked anti-rabbit IgG , HRP-linked anti-mouse IgG , and HRP-linked anti-rabbit IgG from Cell Signaling Technologies . Rac1 pull-down assays were performed using Rac1 activation Assay Biochem Kit ( Cytoskeleton ) following the manufacturer’s instructions . For Calcium flux analysis , CD4 T cells were loaded at 37°C for 30 min with the fluorescent calcium indicator Indo-1 ( Invitrogen ) at 5 μM . The cells were then washed and stained with surface markers ( TCR and CD4 ) . Ca influx was measured by flow cytometry using an LSRII ( BD Bioscience ) . Cells were incubated with 0 . 5 μg/ml of biotylinated anti-CD3 mAb ( 145-2C11 ) at 37°C for 30 sec . Medium were then added and baseline level was measured for 30s , at which time streptavidin ( Sigma-Aldrich ) was added at 1 mg/ml as a cross-linker . Finally ionomycin was added at 2 μg/ml to all samples to verify Indo-1 labeling . The relative concentration of Ca was measured as the ratio between the Ca-bound dye and the Ca free dye . Data were analyzed using FlowJo software . Data are expressed as mean ± s . e . m . The GraphPad Instat statistical package was used for statistical analyses ( GraphPad Software , Inc . , La Jolla , CA USA ) . Results were compared using Mann–Whitney test . Results were considered statistically significant when the p value was <0 . 05 . *: p<0 . 05; **: p<0 . 01; ***: p<0 . 001 . | The understanding of the physiological role of Vav1 , a key regulator of T cell receptor signaling , was primarily inferred from studies using Vav1-deficient mice . Such models , however , provide little insight on how polymorphisms leading to quantitative changes in Vav1 activity could affect immune system functions . In the present study , we focused on a recently identified Vav1R63W natural variant that has been supposed to play a central role in the susceptibility to neuroinflammation . Using a Vav1R63W knock-in mouse model , we show that Vav1R63W leads to defects in adaptor functions and reduces the susceptibility to experimental autoimmune encephalomyelitis , together with an intrinsic defect in the production of Th1/Th17 cytokines by autoreactive effector CD4 T cells . Thus , our study highlights the importance of Vav1 adaptor functions in CD4 T cells differentiation and suggests that genetic or acquired alterations of this Vav1 function could play a major role in susceptibility to Th1/Th17 mediated diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"t",
"helper",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"immunology",
"animal",
"models",
"developmental",
"biology",
"model",
"organisms",
"stem",
"cells",
"molecular",
"development",
"tcr",
"signaling",
"cascade",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"white",
"blood",
"cells",
"animal",
"cells",
"t",
"cells",
"mouse",
"models",
"hematopoietic",
"progenitor",
"cells",
"immune",
"system",
"thymocytes",
"signal",
"transduction",
"cell",
"staining",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"regulatory",
"t",
"cells",
"cell",
"signaling",
"signaling",
"cascades"
] | 2016 | A Natural Variant of the T Cell Receptor-Signaling Molecule Vav1 Reduces Both Effector T Cell Functions and Susceptibility to Neuroinflammation |
Scaling up of insecticide treated nets has contributed to a substantial malaria decline . However , some malaria vectors , and most arbovirus vectors , bite outdoors and in the early evening . Therefore , topically applied insect repellents may provide crucial additional protection against mosquito-borne pathogens . Among topical repellents , DEET is the most commonly used , followed by others such as picaridin . The protective efficacy of two formulated picaridin repellents against mosquito bites , including arbovirus and malaria vectors , was evaluated in a field study in Cambodia . Over a period of two years , human landing collections were performed on repellent treated persons , with rotation to account for the effect of collection place , time and individual collector . Based on a total of 4996 mosquitoes collected on negative control persons , the overall five hour protection rate was 97 . 4% [95%CI: 97 . 1–97 . 8%] , not decreasing over time . Picaridin 20% performed equally well as DEET 20% and better than picaridin 10% . Repellents performed better against Mansonia and Culex spp . as compared to aedines and anophelines . A lower performance was observed against Aedes albopictus as compared to Aedes aegypti , and against Anopheles barbirostris as compared to several vector species . Parity rates were higher in vectors collected on repellent treated person as compared to control persons . As such , field evaluation shows that repellents can provide additional personal protection against early and outdoor biting malaria and arbovirus vectors , with excellent protection up to five hours after application . The heterogeneity in repellent sensitivity between mosquito genera and vector species could however impact the efficacy of repellents in public health programs . Considering its excellent performance and potential to protect against early and outdoor biting vectors , as well as its higher acceptability as compared to DEET , picaridin is an appropriate product to evaluate the epidemiological impact of large scale use of topical repellents on arthropod borne diseases .
Vector-borne diseases remain major contributors to the burden of diseases in the tropics [1] , [2] . The most important vectors for transmission of diseases are bloodsucking arthropods , and especially mosquitoes . Worldwide , about 3500 mosquito species have been described , but only a few of them are able to transmit human disease . The mosquito-borne diseases of public health importance include malaria , filariasis , and arboviral diseases such as dengue , chikungunya , Japanese encephalitis , and yellow fever [1] , [3] . For these diseases , targeting the mosquito instead of the pathogen contributes greatly to disease prevention . Current vector control programs are primarily based on insecticides [1] , [4] . For malaria , which is one of the most serious vector-borne diseases affecting millions of people , upscaling of vector control programs has greatly contributed to its worldwide decrease , and especially in Southeast Asia substantial progresses have been observed [5] . The present vector control programs are primarily based on the distribution of long-lasting insecticidal nets ( LLINs ) and/or application of indoor residual spraying ( IRS ) . However IRS has little impact on outdoor resting vectors , and outdoor and/or early biting species are not affected by LLINs [4] . Some vector species , such as Anopheles arabiensis in Africa [6] , Anopheles maculatus and Anopheles dirus in Asia [7] , [8] , or Aedes aegypti and Aedes albopictus are then less or not vulnerable to one of these two preventive methods . As such , in Southeast Asia , residual malaria transmission due to outdoor and early biting malaria vectors constitutes an important , but often neglected , public health concern in some provinces of each country [9] . Vector control is also of high importance in preventing arboviruses such as dengue ( Flaviviridae ) and chikungunya ( Togaviridae ) as no treatment or vaccine is available [10]–[12] . However , both viruses are transmitted by the day-and outdoor-biting mosquitoes Ae . aegypti and Ae . albopictus [3] , [13] . As early and outdoor biting proportions of vectors will maintain malaria and arbovirus transmission , there is an urgent need for additional control measures tackling these fractions of the vector population [4] . Synthetic repellents are a common means of personal protection against mosquito bites . N , N-diethyl-3-methylbenzamide ( DEET ) is the most commonly used active ingredient in commercially available repellents and has gained wide acceptance in the western world [14] . Another promising synthetic repellent , which was developed by Bayer in the 1980s using molecular modelling , is 1-piperidinecarboxylic acid , 2- ( 2-hydroxyethyl ) - , 1-methylpropylester ( commonly known as picaridin ) . In contrast to DEET , picaridin does not dissolve plastics and other synthetics ( coatings , sealants ) , and is biodegradable . Moreover it is cosmetically more acceptable ( skin feeling , odour ) than DEET [in 14] . The effectiveness of this repellents has shown to equal DEET [15]–[20] , or be better than DEET [21] . Different studies demonstrate the efficacy of topical repellents as personal protection tool against malaria [22]–[26] , whereas others fail to prove such an effect [27] . There is currently no evidence available for repellents as a community protection tool that decreases transmission . The epidemiological efficacy and the impact of topical repellents on malaria and arbovirus transmission will depend on two major factors , which are the performance of the repellent to protect an individual from getting bitten by a mosquito , and the adherence/coverage to repellent treatment in the study community [28] . As such , for implementing the use of repellents in malaria and arbovirus control programs , knowledge on the entomological efficacy of specific repellents is a prerequisite . In Cambodia , a large scale study to raise evidence on the effectiveness of mass use of effective and safe repellents ( picaridin ) in addition to insecticide impregnated bed nets in controlling malaria and arbovirus infections was conducted ( MalaResT project , trial registered as NCT01663831 ) . Present study explores in field conditions the protective efficacy of two formulations of picaridin against the bites of Southeast Asian mosquitoes . A protocol adapted from the World Health Organisation Pesticides Evaluation Scheme guidelines for efficacy testing of mosquito repellents for human skin [29] was applied in which human landing collections were carried out on volunteers applying either a placebo or a test repellent on their exposed limbs . The efficacy and performance of the two formulations of picaridin ( lotion 10% and spray 20% ) were assessed and compared to an ethanol solution of 20% DEET over a period of two years .
The study was carried out in two malaria endemic provinces in Cambodia , namely Mondolkiri ( two villages: Krang Tes ( latitude 12 . 636354N , longitude 107 . 348258E ) and Pou Siam ( latitude 12 . 340183N , longitude 107 . 148045E ) ) and Pailin ( 1 village: Kngok ( latitude 12 . 919693N , longitude 102 . 676803E ) ) , that were chosen based on previous knowledge on the presence of An . dirus s . l . or Anopheles minimus s . l . . As no An . minimus s . l . were collected in Pou Siam , collections were stopped in this village after two surveys , and this village was replaced by Kngok ( Pailin ) . A total of 8 surveys were organized during which mosquito collections took place during 10 days . Pou Siam was only included in surveys 1 and 2 , and Kngok in surveys 3 to 7 , whereas Krang Tes was included in all surveys . In Krang Tes , the study setup was duplicated as from survey 3 onwards . The surveys took place in May , July , September , and November of 2012 and 2013 . In each of the villages , 5 outdoor collection points , near to houses , were chosen which were at least 20 meters apart to avoid mosquito diversion between treated and negative control persons [30] . The protocol of the study was adapted from the WHOPES guidelines for efficacy testing of mosquito repellents on human skin [29] . Five treatments were included in the study: two negative controls ( ethanol ) , one technical grade DEET treatment used as a positive control , given that this repellent is considered as the golden standard ( from Acros Organics diluted at 20% in ethanol ) , and two formulations of picaridin ( 10% repellent lotion and 20% repellent spray formulated by S . C . Johnson ) . The picaridin formulated products complied with the WHO specifications ( confirmed by the chemical analysis at CRA-W , Gembloux , certificate of analysis ITM/FO 23005/Ch . 5362 to 5365/2012/A ) . An experimental replicate consisted of 5 consecutive days during which the lower limbs of 5 persons were treated with repellents or ethanol , followed by mosquito collections on the treated limbs during 5 consecutive hours . This experimental replicate was repeated 46 times over the 8 surveys . The effects of day of treatment , collection site and test person were accounted for by following a 5×5×5 Graeco-Latin Square rotation design . Each day , one of the 5 test persons was assigned to one of the treatments , and the collection sites were rotated among the test persons each hour . Before application of the treatments , the legs of the test person were washed with unscented soap , followed by rinses with clean water and ethanol . The treatments were applied on both legs , between ankle and knee at 1 ml/600 cm2 . Test persons wore long-sleeved shirt , long trousers , and socks up to the ankle . The legs of the trousers were rolled up to the knee to expose only the treated part of the legs to biting mosquitoes . After finishing the test session , the limbs were washed again . Human landing collections were performed starting 30 minutes after treating the legs of 5 trained volunteers , between 17 h and 22 h , except for the last survey , during which collections took place between 19 and 24 h ( but also 30 minutes after treatment of legs ) . There was a continuous exposure to mosquitoes , with a break of 15 minutes at the end of each hour so as to allow the test persons to rest and change collection site . Specimens were collected in labelled individual glass tubes and identified in the field at species level based on morphological characters using identification keys as described in [8] . For An . dirus s . l . , An . minimus s . l . , An . maculatus s . l . , Anopheles barbirostris s . l . , Ae . aegypti , and Ae . albopictus , the parity was determined by examination of the tracheoles within the ovaries in the field [31] . For long-term storage , all mosquitoes were kept dry , in an individual plastic capsule by specimen with the corresponding label . Head and thorax of all anophelines were analysed by the ELISA method for detection of the circumsporozoite protein ( CSP ) as described in [32] . All ELISA positive specimens were subjected to a Plasmodium specific PCR [32] , as false positivity was previously observed in this region . Molecular species identification was performed for mosquitoes morphologically identified as An . dirus s . l . , An . minimus s . l . and An . maculatus s . l . as described previously [8] . All data were collected on standard forms , and were double-entered in a pre-tested Access database by two independent data entry clerks . Databases were compared by using Epi Info ™ 3 . 5 . 3 , and inconsistencies were checked with the hard copy forms and corrected . Repellent efficacy was calculated as percent repellency ( %R ) according to the formula %R = ( ( C-T ) /C ) *100 , Where C is the average of the total number of mosquitoes biting on the lower legs of the two individuals with the control treatment , and T is the total number of mosquitoes biting on the lower legs of a repellent-treated subject [29] . Confidence limits of proportions were calculated according to the Wilson procedure without correction for continuity as described in [33] . Generalized Linear Mixed Models using poisson or negative binomial distributions [34] and their zero-inflated variants ( glmmADMB function in the glmmADMB package applied in R version 3 . 1 . 0 ) were fitted to the data with the daily mosquito count on the treated persons for the different treatments as dependent variable , the treatment and the mosquito genus or vector species and their interaction as explanatory variables , and survey , village , collection day , location , and collector as random factors . Mosquito counts on the treated persons were corrected for the total amount of mosquitoes collected per genus or species on the negative control persons by using the logarithm of the latter as offset in the model . Model comparison was performed by likelihood ratio tests . The final model used a negative binomial distribution , including the treatment and genus/species as fixed effects ( without their interaction ) , and the survey , village and location ( nested within village ) as random effects . Incidence Rate Ratios ( IRR ) were calculated by exponentiation of the model coefficients and their 95% confidence interval . For estimation of the Median Complete Protection Time of each repellent , Kaplan-Meier survival analysis was carried out for each mosquito genus and selected vector species according to [29] . For this analysis , based on the complete protection times ( i . e . time until which one bite was obtained ) recorded per treatment each day , only days during which individuals of the respective genus or species were collected on the negative control persons were included . For studying whether the percent repellency decreased over the five hours of collection , a Chi square for linear trend analysis was performed on the hourly aggregated data per genus or species for each repellent , by using the StatCalc function Chi Square for Trend in Epi Info 7 . The Bonferroni correction was used to correct for multiple comparisons . A logistic regression model was carried out ( glm function in the stats package applied in R version 3 . 1 . 0 ) to study differences in parity rate between treatments and vector species . The model included the parity status of an individual mosquito as outcome ( 0 for nulliparous and 1 for parous ) , and treatment , vector species and their interaction as explanatory variables . Odds Ratios were calculated by exponentiation of the model coefficients and their 95% confidence interval . The study protocol was approved by the ethical committees of the National Centre of Malariology CNM in Phnom Penh ( Cambodia ) and of the University of Antwerp/the Institute of Tropical Medicine of Antwerp ( Belgium ) under Belgian registration number B300201112714 . The mosquito collectors were informed about the objectives , process and procedures of the study and written informed consent was obtained from them . Collector candidates were invited among the adult village population and if individuals wanted to withdraw they were allowed to do so at any time without prejudice . A Rapid Diagnostic Test for malaria diagnosis was done before the start and approximately 14 days after the end of each survey . When required , medical care was provided throughout the study .
In 460 man collection evenings , a total of 5048 mosquitoes were collected on negative control persons , of which 2133 were Culex spp . , 1169 were Mansonia spp . , 664 were Aedes spp . , and 1082 were Anopheles spp . Only Aedes spp . and Anopheles spp . were morphologically identified to species level ( Table 1 ) . Given the low number of mosquitoes collected in Pou Siam , this village was excluded from further analysis . For mosquitoes collected between 5 and 10 pm , biting peaks differed between mosquito genera , being 6–7 PM for Aedes spp . , Culex spp . , and Mansonia spp . , and a steady , slightly rising man biting rate for Anopheles spp . from 6 to 10PM ( Fig . 1A ) . Main vector species Ae . albopictus ( n = 221 ) , Ae . aegypti ( n = 341 ) , An . dirus s . s . ( n = 61 , molecularly confirmed ) , An . minimus s . s . ( n = 247 , molecularly confirmed ) , An . maculatus s . l . ( molecularly confirmed to contain An . maculatus s . s . ( n = 48 ) and An . sawadwongporni ( n = 169 ) ) , and Anopheles barbirostris s . l . ( n = 95 ) were caught in sufficient numbers for the following analyses . Between 5 and 10 PM , biting peaks differed between vectors species , being 6–7 PM for both Ae . albopictus and Ae . aegypti , 7–10 PM for An . sawadwongporni , 9–10 PM for An . minimus s . s . and An . barbirostris s . l . , and a slightly increasing biting rate between 6 and 9PM for An . dirus s . s . and An . maculatus s . s ( Fig . 1B ) . Median complete protection times were calculated to be over five hours for all mosquito genera and all vector species using Kaplan-Meier survival analysis , and could thus not be estimated as the experiment only measured repellent effectiveness for up to five hours . No significant decrease in protective efficacy was observed for the mosquito genera or vector species within the five hours of collection ( S1 and S2 Figs . ) . Repellent performance measured over five hours was generally high , with for all mosquito genera more than 90% of the mosquito bites prevented ( S3a Fig . , Table 2 ) . Picaridin 20% ( %R = 98 . 36% [95%CI: 97 . 78–98 . 79] ) and DEET 20% ( %R = 98 . 60 [95%CI: 98 . 06–98 . 99] ) performed equally well ( IRR 0 . 801 , p = 0 . 517 ) , but more mosquitoes were repelled by DEET & picaridin 20% as compared to picaridin 10% ( %R = 95 . 36% [95%CI: 94 . 46–96 . 12] ) ( p<0 . 01 for both ) . This was the case for all genera , as including the interaction between treatment and genus did not improve the negative binomial model . Independent of the treatment , mosquito repellents were more effective against Mansonia spp . ( %R = 98 . 00 [95%CI: 97 . 22–98 . 57] ) and Culex spp . ( %R = 98 . 19 [95%CI: 97 . 67–98 . 60] ) as compared to Anopheles spp . ( %R = 95 . 92 [95%CI: 94 . 84–96 . 78] and Aedes spp . ( %R = 96 . 53% [95%CI: 95 . 19–97 . 51] ) ( Tables 1 and 3; S3a Fig . ) . Also for the vector species , the repellents performed very well , with at least 90% of the mosquitoes repelled by the repellents with higher concentration of active ingredients ( DEET 20% and picaridin 20% ) , except for An . barbirostris of which only 78 . 95% [95%CI: 65 . 09–88 . 01% ) were repelled by picaridin 20% ( S3b Fig . ) . When modelling the protective efficacy of the repellents only for the selected vector species , similar results were observed for the comparison between repellents as for all mosquito genera: DEET 20% and picaridin 20% exhibited a higher protective efficacy ( %R = 98 . 31% [95%CI: 96 . 91–99 . 08%] and 96 . 44% [95% CI: 94 . 62–97 . 66%] respectively ) as compared to picaridin 10% ( %R = 92 . 37% [95%CI: 89 . 94–94 . 25%] ) , and the interaction between treatment and species did not improve the model . Vector species reacted differently to the repellent treated persons ( Tables 1 , 2 , 4 ) , with Ae . aegypti ( %R = 99 . 41% [95%CI: 98 . 29–99 . 8%] ) and An . minimus s . s . ( %R = 97 . 57% [95%CI: 95 . 45–98 . 72%] ) being more repelled as compared to An . dirus s . s . ( %R = 92 . 35% [95%CI: 85 . 12–96 . 26%] ) , Ae . albopictus ( %R = 94 . 24% [95%CI: 91 . 18–96 . 28%] ) , and An . barbirostris s . l . ( %R = 81 . 75% [95%CI: 74 . 69–87 . 28] ) . As An . maculatus s . s . ( %R = 100% [95%CI: 94 . 87–100%] ) was only collected on the negative control persons , and the model did not converge due to this event , this species was deleted from the analysis . A total of 1040 mosquitoes were processed to define their parity status . The majority of dissected mosquitoes collected on the repellent treated persons were parous ( 66 parous out of the 71 ( 93% ) dissected mosquitoes collected on repellent treated persons , versus 757 parous out of the 969 ( 78% ) dissected mosquitoes collected on the control persons; p = 0 . 014 for pooled mosquito collections on all repellents; Table 5 ) . Although parity rate differed significantly between the vector species ( data not shown ) , no interaction was observed between species and treatment ( p = 0 . 982 ) . All of the anopheline mosquitoes were tested for the presence of Plasmodium falciparum ( PF ) or P . vivax ( PV ) sporozoites by sporozoite ELISA . None of the ELISA positive mosquitoes ( 10 An . hyrcanus for PV210 , 2 An . hyrcanus for PF , 1 An . maculatus s . s . for PV210 ) were confirmed by PCR .
The present study is to our knowledge the most extensive study in Southeast Asia that measures the performance of picaridin repellents on wild anopheline and aedine vectors of malaria and arboviruses . The study was designed to measure the performance of the repellents over a five hour window only , as it was part of a project that measures the epidemiological impact of repellent use on malaria and arboviruses , additional to the use of ITNs . As such , it is important that the current gap in protection [4] due to early and outdoor biting vectors is filled . In general , the repellents tested in this study performed very well , preventing more than 90% of mosquito bites on treated limbs , and with a median Complete Protection Time exceeding the five hours tested in this study . Beside coverage and regular compliance with treatment , repellent performance is an essential parameter for achieving an epidemiological impact on vector borne diseases . Based on a model [28] , in low transmission or pre-elimination areas where most malaria transmission is residual , repellents with 90% entomological efficacy should reduce outdoor malaria transmission by up to 90% when used at a 100% compliance . Even if only about 50% of people comply with the regular treatment of an effective repellent , an additional reduction in transmission of 45% could be obtained . However this model does not consider a possible diversion of mosquitoes to non-repellent compliers [35] . In the present study , two repellent formulations were tested containing different concentrations of picaridin . The spray formulation , which has been shown to be the preferred repellent formulation by adults [36] , contained 20% picaridin , which is considered a safe concentration for long-term use by adults [37] . The 10% picaridin lotion is better suited for application on children as the risk of spraying on sensitive areas of the body ( e . g . eyes , nose , mouth , skin abrasions ) is reduced [38] , and the concentration is adapted to long-term use on children [37] . No significant difference in protective efficacy was observed between an ethanol solution of 20% DEET and the formulated 20% picaridin spray . The formulated 10% picaridin lotion was significantly less effective , although still more than 90% of mosquito bites were avoided . This confirms the findings of equal efficacy of ethanolic solutions of picaridin and DEET against anophelines and aedines obtained in laboratory tests [15] , even if in the current study a commercially available picaridin formulation was compared to the ethanolic DEET solution . Also other field studies find similar protection rates for DEET and picaridin against several mosquito species in Malaysia [16] , [17] , Senegal [18] , Australia [19] , and the USA [20] . In contrast , a field study in Burkina Faso has shown that picaridin has a higher protection rate against several anophelines as compared to DEET [21] . The difference in findings between the current study and the study in Burkina Faso might be due to several factors . First , Cambodia has a different range of anopheline species as compared to Africa [39] , [40] , which could affect the results of this study , as differences in repellent sensitivity were observed between species ( see further ) . Second , in the current study mosquito collections were only conducted during five hours after the application of the repellent . In the above mentioned study on African Anopheles vectors , picaridin always obtained the highest protection as compared to DEET at the end of the 10 hour exposure period [21] . The authors [21] also observed that picaridin remained on the treated limbs longer than DEET , suggesting that the longer-lasting protective efficacy observed with picaridin was presumably not due to higher sensitivity of An . gambiae s . l . to this compound , but rather to a longer residual effect on the skin . It has been shown that moderate levels of physical activity ( jogging , stationary cycling ) can result in a more than 40% decline in complete protection time of some repellents [41] . As such , the longer residual effect , together with the higher acceptance of picaridin as compared to DEET [14] , could make picaridin a more appropriate repellent in vector control programs . Additionally , as no decrease in repellent efficiency over time was observed ( S1 & S2 Figs . ) , repellent sensitivity could be compared between genera and vector species . Differences were observed in the repellent performance between mosquito genera and species . About twice as many Anopheles spp . and Aedes spp . were collected on repellent treated persons as compared to Mansonia spp . and Culex spp . ( Table 3 ) , resulting in a difference in performance ( percent repellency ) of about 2% ( Table 2 ) . Therefore , the present study confirms the findings of laboratory tests , in which picaridin and DEET exhibit higher protection against Culex spp . as compared to aedines and anophelines [42] . In the current study , differences were also observed between vector species , with Ae . aegypti and An . minimus s . s . being the most sensitive to the used repellents , and An . barbirostris s . l . the least , although these differences were less for DEET 20% . Moreover both repellents are more effective against Ae . aegypti than Ae . albopictus . Differences in repellent sensitivity were also observed between closely related species . Although sample size did not allow to detect differences between An . maculatus s . s . and An . sawadwongporni , it is striking that no An . maculatus s . s . were collected on repellent treated persons , whereas repellent insensitivity was observed in An . sawadwongporni . It has been suggested that for field studies the repellent performance cannot be compared between mosquito species [21] , due to decreases in repellent efficiency over time , and concurrent differences in biting peaks or biting densities between species . As such , measuring the effective dose for each species in experimental conditions [43] would provide a more precise estimate for comparing the sensitivity between mosquito populations , but the number of mosquito species available in insectary colonies are limited . Moreover , each mosquito colony passes a bottleneck when established in the laboratory , resulting in degeneration of the gene pool and loss or changes within its behavioural repertoire [44] , making colonized mosquitoes not representative of field populations . Therefore , field studies can provide additional information . As mentioned above , in the current field study no decrease in repellent efficiency was observed over the five hour experiment . The differences observed in performance between mosquito genera or species were therefore not likely to be due to differences in biting times or biting densities . This is illustrated by the fact that the repellents performed better against Culex spp . , with the highest biting densities until five hours after repellent application , as compared to Aedes spp . , with lower biting densities and an early biting peak . Also , Ae . aegypti and Ae . albopictus had similar biting dynamics ( Fig . 1 ) , but differed in their repellent sensitivity . Further , the greatest repellent insensitivity was observed in An . barbirostris with a biting activity which was almost constant between 2 and 5 hours after treatment , and which was only present at low densities . It has been suggested that feeding avidity can also influence repellent insensitivity [43] , [45] . As such vectors with a more anthropophilic trend might exhibit higher repellent insensitivity . In the current study , this is indeed the case for the very anthropophilic vector An . dirus . However , An . barbirostris , which is usually considered a more zoophilic mosquito [46] , showed the highest repellent insensitivity , suggesting other mechanisms , e . g . molecular variations in odour receptors targeted by the repellents . Repellent insensitivity in certain species has indeed been observed in previous studies [17] , [21] , [43] , and can be selected in the laboratory as shown experimentally for Ae . aegypti [47] , and for An . dirus of which a colony established from Chonburi ( Thailand ) in 1968 was tolerant to DEET-concentrations lower than 30% [48] . Unfortunately , the exact mode of action and molecular targets of DEET and picaridin are not yet completely understood , so molecular explanations for the observed genus- and species-specific differences in repellent sensitivity cannot be provided . DEET and picaridin are believed to have an effect on the olfactory system consisting of odorant receptors ( ORs ) that need a common co-receptor ( ORCO ) , and of ionotropic receptors ( IR ) [49] . Recent data support the hypothesis that DEET alters the fine-tuning of the insect olfactory system [50] , as well as triggers a direct response of ORs [51] , [52] , ORCO [52]–[54] or IRs [55] . ORCO and IR40a orthologues are conserved across many insect species , possibly explaining the wide action of DEET as repellent for many insect species [55] , [56] . Few research has been carried out on the mode of action of picaridin , but it has been suggested that picaridin might also target the co-receptor ORCO [57] . Further research to detect genetic alterations ( e . g . mutation , duplication , upregulation ) in these receptors between the currently collected sensitive and insensitive mosquitoes as such could provide key knowledge on the mode of action of both repellents . It has been suggested that the infection status of a mosquito can alter its blood feeding behaviour [58] , and that pathogen infected mosquitoes might respond differently to repellents [59] . In this study however , no malaria sporozoites were detected in any of the collected mosquitoes , which is not surprising regarding the low malaria endemicity ( <5% ) . In previous field studies no significant differences were found in the proportion of anophelines harbouring Plasmodium sporozoites landing on control or repellent treated individuals [21] , [60] . Experimental infections with the four serotypes of Dengue Virus did not alter the responses to DEET of Ae . aegypti and Ae . albopictus [61] , although experimental disseminated Sindbis Virus infection in Ae . aegypti did significantly reduce its time to first bite on DEET and picaridin treated artificial blood meal substrates [59] . As such , until the latter finding is confirmed in the field , it can be assumed that a repellent reducing the number of vector bites , will also reduce the number of infectious bites . Surprisingly , in the present study a higher proportion of parous mosquitoes landed on repellent treated legs as compared to control persons , and this for all vector species involved . This might be related to differences in host avidity between parous and nulliparous mosquitoes as experimentally shown for Ae . albopictus [45] . Despite the statistical analysis being based on a low number of repellent insensitive mosquitoes , it is worth mentioning as the vectorial capacity for a population of vectors is highly dependent on its age structure [62] . As such , older ( parous ) mosquitoes are more likely to harbour infectious pathogens given the extrinsic incubation period of the pathogens in the vector . Therefore , parity status of vector populations should be systematically documented in future field evaluations of repellents and other vector control tools . In conclusion , field evaluation of formulated picaridin repellents shows that the 20% spray formulation performs equally well as the 20% DEET solution , both protecting users from more than 98% of the mosquito bites in the study area . Over the five hour test period , no significant decline in the repellents' efficacy was observed , showing that these repellents can be used as additional personal protection tools against early and outdoor biting vectors . The heterogeneity in repellent sensitivity between mosquito genera and vector species could however impact the efficacy of repellents in public health programs . Considering its excellent performance and potential to protect against early and outdoor biting vectors , as well as its higher acceptability as compared to DEET , picaridin is an appropriate product to evaluate the epidemiological impact of large scale use of topical repellents on arthropod borne diseases . | Malaria and arboviruses are transmitted by several mosquitoes . Targeting these mosquitoes instead of the pathogens can contribute to prevention of these diseases . For mosquitoes biting throughout the night , mosquito nets ( preferably impregnated with insecticides ) are very effective for mosquito control . However , bites of day- , evening- and outdoor-biting mosquitoes have to be prevented in different ways , for example by applying repellents on the skin which contain DEET or other active ingredients such as picaridin . Here we report on the evaluation of the performance of two formulated picaridin repellents ( lotion 10% and spray 20% ) against mosquito bites , including vectors of arboviruses and malaria in the field in Cambodia . These repellent formulations were compared to a DEET solution ( 20% ) . In general , all repellents performed very well , providing more than 97% protection against mosquito bites when used for five consecutive hours . At the highest concentration , the picaridin repellent performed similarly to DEET . However , different mosquito species reacted differently to the repellents . As such , repellents can provide an additional protection against bites of malaria and arbovirus vectors . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"dengue",
"fever",
"neglected",
"tropical",
"diseases",
"tropical",
"diseases",
"viral",
"diseases"
] | 2014 | Field Evaluation of Picaridin Repellents Reveals Differences in Repellent Sensitivity between Southeast Asian Vectors of Malaria and Arboviruses |
Trypanosoma cruzi is the causative agent of Chagas disease , which affects more than 9 million people in Latin America . We have generated a draft genome sequence of the TcI strain Sylvio X10/1 and compared it to the TcVI reference strain CL Brener to identify lineage-specific features . We found virtually no differences in the core gene content of CL Brener and Sylvio X10/1 by presence/absence analysis , but 6 open reading frames from CL Brener were missing in Sylvio X10/1 . Several multicopy gene families , including DGF , mucin , MASP and GP63 were found to contain substantially fewer genes in Sylvio X10/1 , based on sequence read estimations . 1 , 861 small insertion-deletion events and 77 , 349 nucleotide differences , 23% of which were non-synonymous and associated with radical amino acid changes , further distinguish these two genomes . There were 336 genes indicated as under positive selection , 145 unique to T . cruzi in comparison to T . brucei and Leishmania . This study provides a framework for further comparative analyses of two major T . cruzi lineages and also highlights the need for sequencing more strains to understand fully the genomic composition of this parasite .
The protozoan parasite Trypanosoma cruzi , causative agent of Chagas disease , infects 7 . 7 million people in Latin America and causes 12 , 500 deaths annually [1] . Transmission of the parasite most commonly occurs if infected faeces of the haematophagous triatomine insect vector makes contact with mucosae or abraded skin . Most morbidity is associated with the chronic stage of the disease , which can take several years to develop . There is no vaccine against T . cruzi infections and drug treatment is restricted to a small number of drugs with insufficient efficacy and potentially harmful side effects . Multiple genotyping strategies support the subdivision of T . cruzi into six major phylogenetic groups , recently renamed discrete typing units ( DTUs ) I-VI by international consensus [2] . DTU distribution can be loosely defined by several parameters including ecology , vector and host preference , geography and disease association [3] , although patchy sampling precludes definitive associations . Likewise , an accumulating number of in vitro and in vivo experiments indicate significant phenotypic variation between T . cruzi strains in terms of physiology , biochemistry and infectivity [4] , [5] , [6] , [7] , [8] , [9] , [10] , [11] , [12] , [13] . Again , however , there are few clear-cut correlations between genetic groups and pathogenic potential and the genetic determinants of such differences remain enigmatic . Genome sequencing can provide crucial data to facilitate such research . TcI is the predominant agent of Chagas disease in the Americas North of the Amazon e . g . [14] [15] [16] , although it is by no means uncommon in patients in other regions ( e . g . [17] ) . In contrast , TcII , TcV and TcVI are the predominant causes of Chagas disease in the Southern Cone countries , where megaoesophagus and megacolon are more common [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] . TcI shows spectacular abundance among wild hosts and vectors throughout the endemic range of T . cruzi , especially , but not exclusively , in association with Didelphis sp . opossums [3] , [26] . Whereas the other strains responsible for most human disease , TcII , V and VI , are rarely isolated from natural reservoirs or triatomines . Indeed , minimal diversity across multiple markers in putative TcII/TcIII hybrids TcV and TcVI , and their widespread southerly distribution , are consistent with a recent origin alongside domestic transmission cycles ( Lewis et al , submitted ) . In phylogenetic terms TcI and TcII are most divergent and nucleotides models estimate their MRCA at 3-16 MYA [27] . Concurrent with substantial intraspecific genetic diversity , Chagas disease is characterized by a highly variable clinical presentation [1] . This has long been assumed to be , at least in part , a product of genetic differences between strains of T . cruzi [15] . However , despite important advances in T . cruzi genotyping [28] [14] and population genetics [29] , [30] , the genomic variation between lineages or individual clones of T . cruzi remains largely unexplored . The haploid genome of T . cruzi CL Brener ( TcVI ) is approximately 55 Mbp in size [31] . Analyses of the sequence revealed a repeat-rich , hybrid genome , with long regions of conserved synteny to Leishmania major [32] and Trypanosoma brucei [33] . A strong signature of the putative TcII/TcIII hybridization that gave rise to TcVI remains . As such , CL Brener predominantly comprises two divergent haplotypes , named Esmeraldo-like ( TcII ) and non-Esmeraldo-like ( TcIII ) ( abbreviated to Esmeraldo and non-Esmeraldo here ) . The hybrid nature and repetitive content of this genome complicated its assembly , leaving the first T . cruzi genome incomplete by comparison to those L . major and T . brucei . A later effort to place the contigs and scaffolds into predicted chromosomes increased the length of scaffolds , although resolution still requires considerable improvement [34] . We considered the sequencing of a smaller , less repetitive , non-hybrid T . cruzi genome to be a sensible approach to improving resolution . Furthermore , an evolutionarily distinct genome , from a DTU with broader host preferences than TcVI , could provide an interesting basis for comparative genomics . Not only are TcI parasites highly divergent from TcVI in ecology and evolution , but typically they have smaller genomes [28] , [35] , [36] , [37] and have relatively low levels of heterozygosity [30] . They are thus the ideal candidate for analysis . Here we describe shot-gun sequencing and partial genome assembly of Sylvio X10/1 , originally isolated in 1983 from a male individual in Pará State , Brazil , suffering from acute Chagas disease [38] . Sylvio X10/1 is a common reference strain of TcI and is frequently used in both in vivo and in vitro experiments [39] [40] [41] [42] . The genomic contigs and sequence reads were subsequently compared to CL Brener . We found that the core gene content of the two T . cruzi lineages is highly similar , but that they harbor large differences in repetitive content and sequence , which may have functional and epidemiological implications .
This Whole Genome Shotgun project has been deposited at DDBJ/EMBL/GenBank under the accession ADWP00000000 . The version described in this paper is the first version , ADWP01000000 . The data will also be available at TriTrypDB [43] . Trypanosoma cruzi Sylvio X10/1 cells were cultured at 28°C in RPMI liquid medium supplemented with 0 . 5% ( w/v ) tryptone , 20 mM HEPES buffer pH 7 . 2 , 30 mM haemin , 10% ( v/v ) heat-inactivated foetal calf serum , 2 mM sodium glutamate , 2 mM sodium pyruvate and 25 µg/ml gentamycin . Genomic DNA was extracted using the Gentra Puregene Tissue Kit ( Qiagen ) . Sequencing was performed using 454 technology ( FLX/Titanium ) and sequence assembly was performed de novo using the CELERA assembler ( v5 . 4 ) [44] . Gene prediction and annotation was performed using GeneMarkS ( v2 . 6p ) [45] and best reciprocal BLAST hit to CL Brener . Annotations were manually inspected by alignment to CL Brener using Promer [46] and the Artemis Comparison Tool [47] . Gene models were manually added if found to be missing . In cases where genes were disrupted by sequencing errors , all fragments of the genes were annotated . Truncated genes located on contig ends were annotated when possible . Individual genes were identified using reciprocal BLASTp and tBLASTn on both assembled and unassembled reads . Alignments were created using ClustalW and used to call strain-specific differences; both nucleotide differences and insertion-deletion ( indel ) events . Calculation of dN/dS was carried out using yn00 ( PAML , v4 . 2 ) [48] . The McDonald-Kreitman test ( MK-test ) , as implemented in BioPerl ( v1 . 6 ) , was used to evaluate protein adaptation [49] , using alignments created by transAlign [50] with T . brucei used as the outgroup . Synonymous sites were assumed to be neutral while non-synonymous sites were assumed to be deleterious , neutral or confer an advantage . Positive selection was assumed to take place if the number of inter-species non-synonymous changes was greater than the intra-species changes . A contingency table and Fisher's exact test was used to test for significance . The neutrality index ( NI = ( Pn/Ps ) / ( Dn/Ds ) ) was used to test the direction of adaptation , which is expected to be 1 under neutrality , >1 for positive selection and <1 for purifying selection . Using NI , the proportion of adaptive substitutions can be estimated as α = 1 - NI . Sequence reads with similarity to known gene families in CL Brener were analyzed . Initially , homologous genes were collapsed into families using the clustering tool cdhit [51] at a 90% identity threshold . Subsequently clusters were subject to multiple alignments with ClustalW . Profile hidden markov models ( pHMM ) were created using hmmbuild ( v3 , with the parameter –symfrac 0 ) , concatenated to a single file and compressed using hmmpress [52] . Sylvio X10/1 and CL Brener reads were translated into the six reading frames and hmmscan ( with the parameters –nobias and –nonull2 ) was used to conduct searches . To make the results comparable to Sylvio X10/1 , Sanger reads from CL Brener were cut into smaller pieces before the HMM search was conducted . We used 454 technology whole genome shot-gun sequencing [53] to produce a partial assembly as well as a read-based analysis of the TcI reference strain Sylvio X10/1 ( TcI ) genome . We then conducted a comparison to the genome of the reference strain CL Brener ( TcVI ) . This has allowed the first genome-scale analysis of genetic diversity in T . cruzi . The architecture of the two genomes was highly similar , composed of large , co-transcribed , gene-dense “core” coding regions , which displayed highly conserved synteny interspersed with regions of repetitive sequence . The draft assembly has good coverage of these gene dense regions , but is more fragmented in repetitive regions due to the technical difficulties associated with accurate assembly of repeat sequences . However , we have complemented this assembly with a read-based analysis . Thus we were able to characterize comparatively the repeated genes in both genomes . The core gene content of the two genomes was virtually the same but we identified abundant nucleotide and amino acid sequence differences . Furthermore , in the comparison between Sylvio X10/1 and CL Brener we found large differences in the proportion of sequence with homology to multigene families . CL Brener was found to have approximately 5 . 9 Mbp more of haploid sequence related to the DGF , RHS , mucin , MASP , GP63 , and transsialidase gene families . The expansion of these gene families underlies most of the genome size difference between Sylvio X10/1 and CL Brener . Genome sequencing of the TcI isolate Sylvio X10/1 was carried out using 454 technology [53] , which generated 582 Mbp sequence data ( nreads = 1 , 688 , 475 , Table 1 , Figure S1A ) , where 79 Mbp ( nreads = 301 , 005 ) corresponded to maxi/mini circles . Sequence assembly resulted in 7092 contigs ( N50 = 5659 bp ) yielding an average coverage of 11x ( Figure S1B ) . Subsequently , contigs from the assembly were aligned to both CL Brener haplotypes [34] which revealed large blocks of synteny , representing the core gene content of these genomes ( i . e . excluding repetitive regions ) . The amount of heterozygosity in the assembly was examined by counting the number of high quality mismatches between aligned reads , which estimated the heterozygosity to be less than 0 . 08% in the core genome . In the coding regions the mean nucleotide identity was higher between Sylvio X10/1 and non-Esmeraldo i . e . TcIII ( 98 . 2% ) than between Sylvio X10/1 and Esmeraldo i . e . TcII ( 97 . 5% ) ( Table 2 , Figure 1 and 2 ) . The mean nucleotide identity between the two CL Brener haplotypes Esmeraldo and non-Esmeraldo was 97 . 8% . This is independent genome-wide evidence of the generally closer phylogenetic relationship between TcI ( Sylvio X10/1 ) and TcIII ( non-Esmeraldo ) than with TcII ( Esmeraldo ) . The divergence between these three T . cruzi lineages is therefore greater than between T . brucei subspecies T . brucei brucei and T . brucei gambiense ( 99 . 2% ) [54] but less than between two representatives of different Leishmania species complexes , L . major and L . infantum ( 94% ) [55] . From the alignments , a total of 77 , 349 putative fixed differences were identified in the coding regions of a total of 5582 genes ( 8 . 6 Mbp of sequence ) . Of these nucleotide differences 52% were synonymous changes , 34% were non-synonymous changes giving rise to chemically similar amino acids and 23% were non-synonymous changes associated with radical amino acid replacement . The average rate of nucleotide differences ( ND ) between Sylvio X10/1 and non-Esmeraldo was 18 ND/kb/gene and compared to Esmeraldo 25 ND/kb/gene ( Figure 2A ) . In comparison , the average ND rate between non-Esmeraldo and Esmeraldo was 22 NT/kb/gene . This large number of nucleotide differences is consistent with independent evolution of the T . cruzi lineages over several million years [27] , presumably due to ecological , geographic , and/or reproductive isolation , limiting homogenising forces that might act between lineages . Some of these changes may be adaptive , although one explanation for the high proportion of radical amino acid replacements might be low rates of sexual recombination in T . cruzi leading to the accumulation of mildly deleterious mutations over time ( Muller's ratchet ) . Experimental phenotypic comparisons and associated in depth annotation of the potential functional implications of such radical amino acid changes may reveal biological consequences . Multiple CL Brener genes originally thought to have a frame shift not observed in Sylvio X10/1 ( n = 169 , Table S1 ) must now also be considered in such comparisons , because our alignments and confirmatory Sanger sequencing revealed they had been mis-assembled and incorrectly annotated as pseudogenes in CL Brener . Nucleotide substitutions between CL Brener and Silvio X10/1 were not the only coding variations present . A search was also conducted to identify indel events . We identified 1861 coding indels dispersed in 1271 genes . The majority ( n = 1350 , 72 . 5% ) were caused by length variation in microsatellite tracts . Indels 3 bp in length were the most common , followed by 6 and 9 bp . Multiple genes with a functional annotation ( i . e . non-hypothetical genes ) were found to contain indels , for example DNA topoisomerase genes , helicase genes , various metabolic genes and chaperones . Several functionally important genes contained relatively large indels , including the DNA repair protein BRCA2 , which was found to contain a 44 codon N-terminal deletion in Sylvio X10/1 spanning amino acids 82–125 . Although this deletion did not directly affect an evolutionarily conserved domain , it may have functional consequences for BRCA2-mediated homologous recombination capacity in this strain . Deletions were slightly more prevalent in Sylvio X10/1 , which could possibly indicate reductive evolution in Sylvio X10/1 , or , conversely , that sequence expansion has generally been more common in CL Brener . Similarly , the number of 195 bp satellite repeats was greater in CL Brener [56] [36] and the sum of total intergenic distances was marginally larger in CL Brener ( Table 1 ) . The overall content of retroelements , LINEs and LTRs , assessed across both genomes using RepeatMasker and conducted using reads , showed little variation ( Table 1 ) . The clear size differences between the CL Brener and Sylvio X10/1 genomes were confirmed at the macro level . The Sylvio X10/1 haploid genome size was estimated to be 44 Mbp , using extrapolation from the combined length of the contigs from the Sylvio X10/1 assembly ( 23 Mb ) and the unassembled data from repetitive regions ( see following sections ) . Our estimate tallies with previous studies that have estimated the Sylvio X10 genome size at about 35–44 Mbp , using pulse-field gel electrophoresis [37] and flow cytometry [28] . This value for haploid genome size is considerably lower than that for CL Brener ( ∼55 Mbp ) [31] . The smaller genome size appears to be a general feature of TcI strains [28] . We found that Sylvio X10/1 and CL Brener have nearly the same core gene complement , including housekeeping genes , structural genes and genes of unknown function . Six annotated open reading frames ( ORFs ) in CL Brener were not found in Sylvio X10/1 ( Table S2 ) . As these ORFs were short ( <350 aa ) and without a functional annotation , it is unclear whether they are expressed at all . We were not able to identify any Sylvio X10/1-specific genes or significantly long ORFs . However , we note that minimal gene differences are also reported between T . brucei subspecies genomes [54] , as well as between those of Leishmania species [55] . A similar trend has been observed in Giardia lamblia [57] , [58] . Instead , the great majority of genetic differences between strains of all these parasite genera consist of SNPs and indels as well as , crucially , copy number ( see following section ) . In the absence of strain specific genes in our dataset , we also screened for those genes that might be under directional selection between Silvio X10/1 and CL Brener . dN/dS ratios ( ω ) identified 336 genes under positive selection ( ω >1 ) , a significant proportion of which ( 145 ) were unique to T . cruzi by comparison to T . brucei and Leishmania . The presence of these rapidly evolving T . cruzi specific genes could indicate important biological roles in American trypanosomes , for example , genes regulating interactions with hosts or vectors . Those genes that could be assigned function included two genes encoding cell-surface targeted proteins , one 90 kDa surface protein gene and one member of the TolT family . MK tests ( see Materials and Methods ) for adaptive selection between T . cruzi and T . brucei identified other genes of known function and putative importance including transporters and various other membrane coupled proteins , as well as , surprisingly , some DNA repair proteins , chaperones and cyclins ( Table S2 ) . Many surface proteins involved in interaction with the host in T . cruzi are encoded by large repetitive gene families [31] . These regions represent a major area of interest for comparisons between CL Brener and Sylvio X10/1 genomes . Assembly of such repetitive sequences is problematic , therefore we applied a novel approach . The Sylvio X10/1 assembly contained only about 49% of the generated sequence data , leaving 710 , 109 reads ( ∼236 Mbp ) that did not enter the assembly . To evaluate these data , sequence reads were classified into pre-defined categories using profile hidden markov models . The size of each gene family was estimated using the combined alignment length and normalized to the total amount of sequence data ( Figure 3 ) . To provide an estimate of the relative repeat abundance , the same searches were performed on the CL Brener sequence data . To verify the applied method , several single copy genes were included in the analysis . The vast majority of the expected single copy genes resulted in a 1∶1 signal , indicating that the method can be used reliably for copy number quantification . By this classification approach , a total of 346 , 696 ( 49% , 137 Mbp ) unused reads from Sylvio X10/1 were sorted into 69 different categories ( Figure 3 ) . From these unused reads , 233 , 574 ( 33% , 92 Mbp ) were assigned to six categories only ( sialidase , DGF , RHS , mucin , MASP and GP63 ) . In terms of combined alignment length , these gene families were estimated to represent 7–8 Mbp of the haploid Sylvio X10/1 genome . For Sylvio X10/1 and CL Brener , the sialidase and DGF categories were the largest for each genome respectively , comprising 5 . 4% and 6 . 1% of the sequence data . According to this analysis , a smaller proportion of the sequence reads match the DGF family in Sylvio X10/1 , suggesting that this family is expanded in CL Brener or contracted in Sylvio X10/1 . The analysis also indicated copy number differences for the MASP , mucin , GP63 and RHS gene families between the two genomes . It should be noted that this method does not discriminate between pseudogenes and functional genes and therefore , some of the predicted genes could represent non-functional or non-expressed gene variants . In addition to inter-genomic comparisons between the major gene families , a more comprehensive analysis was performed on a larger set of T . cruzi genes , which included 5874 different homologous gene clusters , including singletons . The most significant differences were found among some hypothetical genes , and in most cases there was an expansion in CL Brener . These comparative analyses of both the non-coding and coding repetitive elements indicates significant differential expansion in sequence corresponding to surface antigen repertoires and other multicopy gene families . The CL Brener genome was estimated to have about 5 . 9 Mbp ( 11 . 8 Mbp diploid ) of extra sequence related to multigene families than Sylvio X10/1 . Therefore , we conclude that expanded gene families in CL Brener underlie most of the genome size difference between TcI and TcVI , and this may theoretically enhance functional plasticity . CL Brener ( TcVI ) is the product of hybridization between TcII and TcIII [59] . We cannot determine whether the gene family expansions occurred pre- or post-hybridisation ( or both ) . However , TcII , TcIII and TcVI strains all have similarly increased DNA contents relative to TcI [28] . This suggests the bulk of expansion occurred within ancestral TcII and TcIII . This first intra-species comparative genomic analysis of T . cruzi provides several significant insights . First , it is clear that core genome synteny and gene identity are highly conserved between TcI and TcVI , with very few unique and no major gene differences . Similarly , the overall quantity of non-coding DNA is largely unchanged between the two genomes . The most significant variation between the two genomes is in the size of several multigene families , which encode many important surface proteins . These families are significantly larger in TcVI and account for approximately 54% of the c . 11 Mbp size difference between TcVI and TcI . Our findings compare well with recent comparative genomic studies of other parasitic trypanosomes at the sub-species ( T . brucei , [54] ) and species complex ( Leishmania , [55] ) level . In both cases few gene differences are apparent in the core genomes , congruent with the remarkable synteny observed at the inter-species level [31] . This apparent lack of genomic rearrangement , gene deletion and insertion between trypanosome genomes could derive from the constraints of polycistronic transcription , disruptions of these long co-transcribed gene clusters being likely to be deleterious . Genetic recombination is a common mechanism by which structural change may be introduced between genomes , as well are providing sources of new genetic information . The excessive accumulation of non-synonymous changes that we observe between TcI and TcVI suggest that this recombination may be infrequent in T . cruzi at the inter-DTU level at least . However , the overall natural frequency of intra-species and intra-genotype genetic recombination in all three major human parasitic trypanosome genera is a still a matter of some uncertainty and considerable debate [60] , [61] , [62] , [63] , [64] , [65] . Functional dissection of the larger surface gene families in TcVI presents an interesting problem . Both TcI and TcVI efficiently infect humans and TcVI is found among far fewer hosts than TcI [3] . However , TcVI may have emerged quite recently in conjunction establishment in the human host ( Lewis et al , submitted ) . It remains to be defined how much of the differential surface gene diversity is actually expressed . This study represents a significant advance in unraveling the diversity of T . cruzi and encourages further comparative genomics of the T . cruzi lineages and related species of the subgenus Schizotrypanum . We are currently engaged in sequencing other representatives of TcI , and the apparently bat specific trypanosome T . cruzi marinkellei . | Chagas disease is a major health problem in Latin America and it is caused by the protozoan parasite Trypanosoma cruzi . The genome sequence of the T . cruzi strain CL Brener ( TcVI ) has revealed a genome with large repertoires of genes for surface antigens , among other features . In the present study , we sequenced the genome of a representative member of TcI , the predominant agent of Chagas disease North of the Amazon and performed comparative analyses with CL Brener . Genetic variation between strains can potentially explain differences in disease pathogenesis , host preferences and aid the identification of drug targets . Our analysis showed that the two genomes have very similar sets of genes , but contain large differences in the relative size of several important gene families . Moreover , an abundance of allelic sequence variation was found in a large fraction of genes , and an evolutionary analysis indicated that many genes have evolved at different rates . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/tropical",
"and",
"travel-associated",
"diseases",
"genetics",
"and",
"genomics"
] | 2011 | Shotgun Sequencing Analysis of Trypanosoma cruzi I Sylvio X10/1 and Comparison with T. cruzi VI CL Brener |
In Caenorhabditis elegans , the dosage compensation complex ( DCC ) specifically binds to and represses transcription from both X chromosomes in hermaphrodites . The DCC is composed of an X-specific condensin complex that interacts with several proteins . During embryogenesis , DCC starts localizing to the X chromosomes around the 40-cell stage , and is followed by X-enrichment of H4K20me1 between 100-cell to comma stage . Here , we analyzed dosage compensation of the X chromosome between sexes , and the roles of dpy-27 ( condensin subunit ) , dpy-21 ( non-condensin DCC member ) , set-1 ( H4K20 monomethylase ) and set-4 ( H4K20 di-/tri-methylase ) in X chromosome repression using mRNA-seq and ChIP-seq analyses across several developmental time points . We found that the DCC starts repressing the X chromosomes by the 40-cell stage , but X-linked transcript levels remain significantly higher in hermaphrodites compared to males through the comma stage of embryogenesis . Dpy-27 and dpy-21 are required for X chromosome repression throughout development , but particularly in early embryos dpy-27 and dpy-21 mutations produced distinct expression changes , suggesting a DCC independent role for dpy-21 . We previously hypothesized that the DCC increases H4K20me1 by reducing set-4 activity on the X chromosomes . Accordingly , in the set-4 mutant , H4K20me1 increased more from the autosomes compared to the X , equalizing H4K20me1 level between X and autosomes . H4K20me1 increase on the autosomes led to a slight repression , resulting in a relative effect of X derepression . H4K20me1 depletion in the set-1 mutant showed greater X derepression compared to equalization of H4K20me1 levels between X and autosomes in the set-4 mutant , indicating that H4K20me1 level is important , but X to autosomal balance of H4K20me1 contributes only slightly to X-repression . Thus H4K20me1 by itself is not a downstream effector of the DCC . In summary , X chromosome dosage compensation starts in early embryos as the DCC localizes to the X , and is strengthened in later embryogenesis by H4K20me1 .
Dosage compensation equalizes X chromosome gene expression between sexes . Different animals use different strategies of dosage compensation by co-opting diverse mechanisms of gene regulation to the X chromosome [1] . In mammals , dosage compensation transcriptionally inactivates one of the two X chromosomes in XX females to equalize overall X expression to that of XY males . In Drosophila melanogaster , the X chromosome is transcribed two-fold higher in XY males . In Caenorhabditis elegans , dosage compensation represses both X chromosomes by half in XX hermaphrodites , equalizing overall X-chromosomal transcript levels to that of XO males . X chromosome dosage compensation is established during , and is essential for development in mammals , D . melanogaster and C . elegans . In mice , failure to inactive the X results in continual deterioration of the embryo and death around 10 days post coitum [2–4] . In D . melanogaster , the dosage compensation complex ( male specific lethal ( MSL ) complex ) localizes to the X chromosome at the late blastoderm/early gastrula stage [5 , 6] . Mutations in any of the four MSL complex members slow development and lead to lethality at the late larval and early pupal stages [7–9] . In C . elegans , mutations in several dosage compensation complex ( DCC ) subunits are maternal effect lethal , where the progeny of homozygous null mutant worms die at early larval stages [10–12] . Mammalian X inactivation , D . melanogaster MSL complex , and the C . elegans DCC all regulate X chromosome chromatin structure [1 , 13] . In mammals , X inactivation leads to enrichment of various heterochromatic histone modifications on the inactive X [14 , 15] . In D . melanogaster , MSL complex increases H4K16 acetylation on the male X chromosome via its histone acetyl transferase subunit MOF [16 , 17] . In C . elegans , DCC binding is required for H4K20me1 enrichment and H4K16ac depletion on X chromosomes in hermaphrodites [18 , 19] . Although several histone modifications are associated with dosage compensation , it is unclear if these histone modifications act as downstream effectors of the dosage compensation complexes . In C . elegans , the five-subunit core of the DCC is a condensin complex , called condensin IDC ( hereafter condensin DC ) [20] . Condensin DC interacts with additional proteins that have roles in hermaphrodite and X-specific recruitment of the DCC to the X chromosomes [21–24] . Condensins are evolutionarily conserved protein complexes that are essential for chromosome condensation and segregation during cell division ( reviewed in [25] ) . In metazoans , there are two types of condensins , named condensin I and II . In C . elegans , condensin DC is distinguished from condensin I by a single subunit , DPY-27 [20] . Unlike condensin DC , condensin I and II bind to all chromosomes [20 , 26] . In C . elegans , knockdown of condensin II specific subunit KLE-2 suggested that condensin II is also repressive [26] . Condensins have been implicated in transcriptional regulation in other organisms , but the molecular mechanisms by which condensins regulate transcription remain unknown ( reviewed in [27] ) . C . elegans DCC is as a clear paradigm for studying the mechanisms of transcription regulation by condensins . The DCC begins localizing to the hermaphrodite X chromosomes at around 40-cell stage ( Fig 1A ) [22 , 28] . DCC binding leads to enrichment of H4K20me1 on the X that begins around the 100-cell stage [18 , 29] . Here , we analyzed X chromosome dosage compensation during C . elegans development by performing mRNA-seq comparison of gene expression between sexes , and in dpy-27 ( condensin DC subunit ) and dpy-21 ( non-condensin DCC member ) mutants . Our results suggest that condensin DC starts repression as it localizes to the X chromosomes in early embryos ( 4–40 cell stage ) . Gene expression differences between dpy-21 and dpy-27 mutants suggest an additional DCC independent role for dpy-21 , emphasizing that condensin DC core and the non-condensin DCC members functionally differ . To clarify the role of H4K20me1 in X chromosome dosage compensation , we performed mRNA-seq and H4K20me1 ChIP-seq analyses in set-1 ( H4K20 monomethylase ) and set-4 ( H4K20 di-/-tri methylase ) mutant worms . Our results suggest that H4K20me1 levels are important for X chromosome dosage compensation , but H4K20me1 by itself does not act as a downstream effector of the DCC .
We measured X chromosome dosage compensation in early embryos , when zygotic transcription is active but the DCC is not yet fully localized to the X chromosome . Here , we measured dosage compensation of a gene as the ratio of its mRNA levels between hermaphrodites and males . We generated mRNA-seq datasets from 2–5 biological replicates ( S1 File ) that highly correlated ( S1 Fig ) . Because of the difficulty in obtaining pure male populations prior to the onset of sex specific gene expression , we compared hermaphrodite embryos ( XX ) to mixed sex ( XX and XO ) embryos . Mixed sex embryos were isolated by crossing males and hermaphrodites of obligate out-crossing strain ( fog-2 ( oz40 ) ) to ensure a 50% male population . “Early embryos” are between 4-cell and 40-cell stages ( S2 Fig ) . At this stage , zygotic transcription is activated and the DCC starts localizing to but not fully enriched on the X chromosome in all cells ( Fig 1A ) [22 , 29] . In early embryos , X chromosome expression was higher in hermaphrodites compared to mixed sex ( Fig 1B ) . The median log2 ratio of expression between hermaphrodite and mixed sex embryos was significantly higher on the X ( 0 . 182 ) compared to autosomes ( ranges from 0 . 034 to 0 . 052 ) ( Fig 1C , one sided Wilcoxon rank-sum test p < 1 . 77 x 10−9 ) . Additionally , genes with significantly higher expression in hermaphrodites were enriched on the X ( Fig 1D , Fisher’s exact test p = 3 . 12x10-6 ) . Thus , X chromosome is not fully dosage compensated between sexes in early embryos , but X chromosomal transcripts are fairly similar between sexes , as the observed median log2 expression ratio of 0 . 182 is less than expected if X expression were two-fold higher in XX hermaphrodite embryos in the XX/mixed sex comparison ( 0 . 411 ) . One mechanism that might equalize X chromosomal transcript levels between sexes is the maternal loading of mRNAs into oocytes . To test this , we asked if similarly expressed genes are maternally loaded . While most methods identify genes that are differentially expressed between sexes , a lack of differential expression does not show that genes are statistically similarly expressed . We identified 780 genes that are statistically similarly expressed between hermaphrodite and mixed sex early embryos by taking genes that show less than 30% difference in expression within a 95% confidence interval ( see Materials and Methods for full explanation ) . Among the 36 similarly expressed X-chromosomal genes , 35 were maternally loaded ( FPKM>1 in oocytes [30] ) . Approximately 95% of similarly expressed genes on autosomes were also maternally loaded , thus maternal loading of transcripts into oocytes helps equalize gene dosage between hermaphrodite and male embryos ( Fig 1E ) . To specifically study dosage compensation of zygotic gene expression , we identified genes that are not expressed in 2-cell stage embryos but are newly expressed in early embryos . K-means clustering of expression between previously published expression data in 2-cell hermaphrodite embryos [30] and our data in early embryos resulted in five distinct clusters ( Fig 2A ) . Genes in cluster 3 are zygotic because they showed little or no expression in 2-cell embryos and increased expression in early embryos ( Fig 2B ) . Since we cannot distinguish between stable and newly transcribed mRNAs , genes in the remaining clusters may also be zygotically expressed . Thus , to study zygotic expression , we focused on the newly expressed genes ( cluster 3 ) . Hermaphrodite expression from the X chromosome was significantly higher compared to the autosomes ( one sided Wilcoxon rank sum test , p = 0 . 019 ) , but the difference was less than expected if there were no dosage compensation . Therefore , early zygotic expression from the X is dosage compensated , but not completely ( Fig 2C ) . In C . elegans , incomplete dosage compensation in early embryos is logical , as sex is determined by the ratio of X and autosomal sex elements ( XSEs and ASEs ) [31–33] . In early embryos , we observed that the male-specific her-1 gene is expressed 11 . 2 fold higher in mixed-sex samples compared to hermaphrodites ( Fig 2D ) . This corresponds to an estimated 17 fold higher expression in males , consistent with previous measurements of ~20-fold difference in her-1 expression between males and hermaphrodites [34] . Interestingly , male promoting gene xol-1 transcript was only slightly higher in mixed sex samples ( Fig 2D ) . It is possible that the difference in endogenous xol-1 transcript level between sexes is smaller than seen for transgenes [35] . It is also possible that the X-linked xol-1 is being repressed by the DCC in early embryos , as dpy-27 starts to repress the X chromosomes in early embryos ( Fig 2E ) , and xol-1 was shown to be repressed by the DCC [36] . We also found that mRNA levels of both XSEs and ASEs were higher in hermaphrodites , suggesting that XSE and ASE dosage does not function through a simple ratio of mRNA levels between sexes ( S3 Fig ) . Instead , small and transient differences in the transcription of XSEs , ASEs and xol-1 may be amplified post-transcriptionally to ultimately regulate her-1 expression , which is clearly higher in males ( Fig 2D ) . The DCC is coupled to sex determination through sdc-2 , which is specifically expressed in hermaphrodite embryos [22] . SDC-2 is required for DCC localization to the X , and for repressing male-specific her-1 gene [37] . A recent study showed that in 50-cell stage embryos , half the cells contain DPY-27 specifically on the X chromosomes [29] . In our early embryo collections ( 4–40 cell stage ) , sdc-2 is more highly expressed in hermaphrodites and her-1 is expressed in males , therefore the DCC is activated ( Fig 2D ) . To test if the DCC represses X-chromosome expression shortly after localizing to the X , we collected dpy-27 null mutant embryos using a strain with genetically balanced ( y56 ) allele ( Materials and Methods ) . In early embryos , dpy-27 null mutation caused significant derepression of newly transcribed zygotic genes on the X ( Fig 2E ) , suggesting that dpy-27 represses X chromosomes in early embryos . Dpy-27 is not detectable in the XX hermaphrodite germline by immunostaining , but it is maternally deposited and required for development [38–40] . Therefore , we interpret the observed dpy-27 null mutation effect coming from the embryos that loaded the DCC onto the X chromosomes in some cells by the 40-cell stage . This implies that the DCC is able to repress X chromosomes quickly after loading . To determine if and when average X chromosome expression is equalized between sexes , we analyzed further time points including comma stage embryos , L1 and L3 larval stages , and young adults . At the comma stage , X chromosome expression remained slightly but significantly higher in hermaphrodites compared to mixed sex worms ( Fig 3A ) . In , L1 , L3 larvae and young adults , average X-chromosome expression was no longer higher in hermaphrodites . In addition to measuring dosage compensation at each time point , for each gene we also calculated change in dosage compensation between two consecutive time points . From early to comma stage embryos , the X chromosome was slightly repressed , but X chromosome gene expression became significantly more balanced between sexes after the comma stage ( Fig 3B , Wilcoxon rank sum p = 1 . 13 x 10−63 ) . To address if more genes are dosage compensated after the comma stage , we identified the statistically similarly expressed genes at each developmental stage . The percentage of X chromosomal genes that were statistically similarly expressed between sexes increased from early to comma stage embryos ( 4 . 6% to 12 . 8% , respectively ) , and stayed at similar levels in L1 , L3 and young adults ( 10 . 3% , 10 . 4% , 13 . 5% , respectively ) . Thus , on average more X chromosomal genes are repressed after the comma stage ( Fig 3B ) , but the proportion of equivalently expressed X chromosomal genes increases after early embryogenesis . In comma stage embryos , average X expression may be higher in hermaphrodites due to stable transcripts retained from early embryos . To test this , we used K means cluster analysis and further filtering to identify those genes that are newly expressed in comma stage embryos and in L1 larvae ( Fig 3C ) . The newly expressed X chromosomal genes in L1 larvae were dosage compensated and repressed by the DCC ( Figs 3C and S5 ) . In comma stage embryos , the newly expressed X chromosomal genes showed significantly higher hermaphrodite expression compared to autosomal genes , thus transcript stability does not fully explain significant hermaphrodite-biased expression from the X chromosomes . It is possible that in comma embryos , X is expressed higher in hermaphrodites because DCC-mediated dosage compensation is incomplete or more genes on the X chromosome have hermaphrodite-biased expression . Comparison of gene expression between sexes is a measure of dosage compensation , but sex-biased gene expression confounds the interpretation of the data ( see discussion ) . To specifically study DCC mediated X repression , we analyzed gene expression changes in hermaphrodites mutant for dpy-27 or upon dpy-27 RNAi knockdown . As mutants and RNAi treated worms showed more variability in staging , we mainly used “mixed stage embryos” isolated by bleaching gravid adults . Mixed stage embryos contained 100–300 cells ( S4A Fig ) . We used dpy-27 RNAi in mixed embryos and L3 , because of the difficulty collecting dpy-27 ( y56 ) null mutant . Western blot analyses showed ~70% and ~40% knockdown of DPY-27 in embryos and L3s , respectively ( S4B Fig ) . Mutation ( Fig 3D ) or depletion ( Fig 3E ) of dpy-27 caused significant X chromosome derepression in early and mixed-stage embryos , L1 and L3 worms . Analysis of newly expressed genes in each stage also indicated that condensin DC represses X chromosomes throughout these developmental stages ( S5 Fig ) . Therefore , although previous studies that used temperature sensitive mutants of various DCC subunits found that the critical time period for DCC activity is centered around the comma stage of embryogenesis , our results suggest that the DCC acts throughout development [12] . We also analyzed dpy-21 , a non-condensin subunit of the DCC , whose null mutation ( e428 ) is not lethal , but leads to X chromosome dosage compensation defects including a dumpy phenotype [10 , 11 , 41] . In mixed embryos and L3 , dpy-21 ( e428 ) mutation also caused X chromosome derepression ( Fig 3F ) . In early embryos , dpy-21 ( e428 ) had a relatively less X-specific effect compared to dpy-27 ( y56 ) . In dpy-21 ( e428 ) early embryos , 72% of X and 58% of autosomal genes were significantly upregulated . In dpy-27 ( y56 ) early embryos , DESeq analysis did not identify any genes that met the significance cutoff ( adjusted p value < 0 . 05 ) . But , among the top five percent of dpy-27 regulated genes , 81% of X and 31% autosomal genes were upregulated . The effect of dpy-21 ( e428 ) on X chromosome expression was more specific in the later stage worms . Clustering and heatmap analysis of gene expression changes showed that dpy-21 ( e428 ) caused both increased and decreased expression from the X chromosomes in early embryos , whereas majority of the X chromosomal genes were derepressed in comma stage embryos and L3s ( S4C Fig ) . A recent study noted that regulation of H4K16 acetylation on the X chromosomes differs between dpy-21 and other DCC subunits in early embryos [29] . Since regulation of X chromosome expression also differed between dpy-27 ( y56 ) and dpy-21 ( e428 ) mutants , dpy-21 may have a DCC-independent role in early embryos . An earlier microarray study suggested that approximately half of X-chromosomal genes are dosage compensated . These genes were identified by the criteria that they were upregulated in the DCC mutants and were equally expressed between XX and XO hermaphrodite embryos [42] . Subsequently , global run on analysis ( GRO-seq ) of active transcription in sdc-2 ( y93 , RNAi ) embryos showed a uniform increase in RNA Pol II levels across most X chromosomal genes [43] , suggesting that there is no large distinct set of genes that escape from DCC-mediated repression . In agreement with a continuous effect , our analysis of the published GRO-seq data showed similar upregulation between previously categorized dosage compensated and non-compensated genes ( Fig 4A ) . In mixed stage embryos , dpy-27 RNAi mRNA-seq also showed derepression of X chromosomal genes . Derepression was slightly stronger for previously defined compensated genes , suggesting that posttranscriptional regulation contributes to the final mRNA levels in the DCC mutants . To further test if the DCC acts on a distinct set of genes on the X , we analyzed the distribution of expression changes caused by DCC disruption . We used a standard expectation-maximization algorithm that mixes multiple normal distributions to model the overall pattern of expression changes . We reasoned that if there is a large group of genes specifically repressed by the DCC , when two distributions are forced on the data , one distribution should reflect expression of DCC regulated genes and another should reflect genes not regulated by the DCC . Indeed , when applied to the whole genome GRO-seq data , two distributions can be modeled . Separation of two distributions with average fold changes of -0 . 37 and 0 . 69 closely mirrored the distributions of changes for the autosomes and X chromosome ( Fig 4B ) . This indicated that the analysis is sensitive to finding differentially regulated genes by the DCC . Next , we applied the same method to the X chromosome , using GRO-seq data ( Fig 4C ) , and the mRNA-seq data from dpy-27 ( y56 ) L1s ( Fig 4D ) . In both cases , the separated distributions all showed upregulation , suggesting that there are no large group of genes that escape DCC regulation on the X . This does not exclude the possibility that different groups of genes are affected at different levels , but this effect must be continuous , rather than discrete . In C . elegans , H4K20 is monomethylated by SET-1 , and H4K20me1 is converted to di- and tri-methylation by SET-4 [18 , 19] . Concurrent enrichment of H4K20me1 and depletion of H4K20me3 on the X chromosomes suggested that the DCC increases H4K20me1 levels on the X by reducing SET-4 activity [18] ( Fig 5A ) . X-enrichment of H4K20me1 requires the DCC subunits , including dpy-21 [18] . We analyzed genome-wide expression changes that occur in dpy-21 ( e428 ) , set-1 ( tm1821 ) and set-4 ( n4600 ) null mutants [18 , 19] . Set-1 null mutant is maternal effect sterile , and RNAi knockdown of set-1 leads to germline deficient adults , thus we could not isolate embryos . Therefore , we collected set-1 ( tm1821 ) homozygous and heterozygous L3 larvae ( see Material and Methods ) . Western blot analysis in whole-larval extracts showed expected changes in H4K20me1 ( Fig 5B ) [18 , 19] . H4K20me1 reduced in dpy-21 ( e428 ) , increased in set-4 ( n4600 ) , and was eliminated in set-1 ( tm1821 ) . Immunofluorescence analysis of intestinal nuclei had also shown that H4K20me1 is higher on X in N2 wild type worms; decreases to the autosomal levels in dpy-21 ( e428 ) mutant; and increases and becomes equal between X and autosomes in set-4 ( n4600 ) mutant [19] . H4K20me1 reduction due to set-1 or dpy-21 null mutation led to a significant increase in X chromosome expression compared to autosomes in L3 larvae ( Fig 5C ) . We observed a small but significant increase in X expression compared to autosomes in the set-4 ( n4600 ) mutant L3 ( Fig 5C ) and mixed stage embryos ( Fig 5D ) . Expression of previously defined dosage compensated genes [11 , 23 , 42] were increased in dpy-21 ( e428 ) , set-1 ( tm1821 ) , but not as much in set-4 ( n4600 ) mutant ( Fig 5E ) . To test if a subset of X chromosomal genes were responsible for the effect seen in the set-4 ( n4600 ) mutant ( Fig 5C ) , we plotted the distribution of expression ratios on the X and autosomes ( Fig 5F ) . This indicated a slight shift between X and autosomes , suggesting that set-4 ( n4600 ) has a subtle effect across all genes , rather than a large effect on a subset of X chromosomal genes . Note that mRNA-seq measures relative expression changes , rather than absolute . Therefore X derepression seen in the set-4 ( n4600 ) mutant may not be specific derepression of the X , as discussed later in the results . Nevertheless , the effects of set-1 and set-4 mutations indicate that proper regulation of H4K20 methylation is important for X chromosome dosage compensation . GRO-seq analysis of sdc-2 ( y93 , RNAi ) embryos showed that the DCC reduces RNA Pol II at the X chromosome promoters [43] . To test if the expression changes seen in dpy-21 ( e428 ) and set-4 ( n4600 ) mutants are due to transcription , we performed AMA-1 ( RNA Pol II large subunit ) ChIP-seq analysis . Average RNA Pol II enrichment across the transcription start and end sites showed a 3’ accumulation that was also noted in the GRO-seq analysis for C . elegans genes ( Fig 5G ) [43] . RNA Pol II on the X chromosome promoters increased relative to autosomes in dpy-21 ( e428 ) mutant ( Fig 5G ) . In the set-4 ( n4600 ) mutant ( Fig 5G ) , there was a subtle shift in RNA Pol II levels between the X and autosomes , suggesting that the effect of set-4 is also at the level of transcription . These results do not exclude post-transcriptional effects contributing to increased X expression in the dpy-21 ( e428 ) and set-4 ( n4600 ) mutants , but do suggest that dpy-21 and set-4 act at the level of transcription . Unlike other DCC subunits , dpy-21 is not essential , but is required for H4K20me1 enrichment on the X chromosomes [11 , 18 , 19] . To understand the role of dpy-21 in the DCC , we compared gene expression changes in dpy-27 RNAi , dpy-21 ( e428 ) , set-1 ( tm1821 ) and set-4 ( n4600 ) mutant L3 larvae by clustering differential expression ratios on the X chromosome and autosomes ( Fig 6A ) . We note that while the effect of dpy-27 RNAi and dpy-21 ( e428 ) were generally similar on the X , genes in several clusters were affected differently between dpy-27 RNAi and dpy-21 ( e428 ) . We performed gene ontology ( GO ) analyses comparing enriched functions in each cluster compared to the whole genome ( Fig 6B ) . Interestingly , the effect of dpy-27 RNAi and dpy-21 ( e428 ) on clusters enriched for cuticle genes were opposite , yet both mutations result in a dumpy phenotype . Since our model for DCC regulation of H4K20me1 posits that the DCC reduces SET-4 activity , X derepression in the set-4 mutant was somewhat unexpected . To study this further , we performed H4K20me1 ChIP-seq analyses in wild type , dpy-21 ( e428 ) and set-4 ( n4600 ) mutant embryos and L3 larvae . Experimental replicates correlated well with each other ( S6A Fig ) . H4K20me1 is enriched in active gene bodies in multiple species [44] . In C . elegans hermaphrodite L3s , H4K20me1 is enriched on expressed genes , but also across the entire X chromosome including silent genes and intergenic regions ( Fig 7A ) . Such pattern of H4K20me1 enrichment on the X could be achieved by a uniform deposition of H4K20me1 by SET-1 during M phase , followed by DCC-mediated reduction of SET-4 activity across the X chromosome [18] . However , standard ChIP-seq analysis cannot detect a uniform increase , due to the analysis step that normalizes read depth between experiments [45 , 46] . Therefore , to analyze change in H4K20me1 in the absence of SET-4 , we used C . briggsae ChIP extract as a spike-in control , similar to previous approaches ( Fig 7B ) [45 , 47] . The replicates correlated with each other ( S6B Fig ) . Here , we assume that the ChIP-seq enrichment calculated from the C . briggsae reads should be the same between set-4 ( n4600 ) and wild type , since the spiked in C . briggsae extract is the same . For each ChIP sample , the corresponding input was also sequenced , thus the normalization is internally controlled against any potential difference in the spiked in proportion . We normalized H4K20me1 ChIP enrichment from C . elegans reads to that of C . briggsae reads , and calculated the ratio of H4K20me1 levels between set-4 ( n4600 ) and wild type worms ( Fig 7B ) . In set-4 ( n4600 ) L3s , there was ~10-fold increase in H4K20me1 on autosomes and ~6 fold increase on the X chromosome ( Fig 7C ) . This is consistent with the idea that in the wild type worms , the DCC already reduces SET-4 activity on the X by half , thus in the absence of SET-4 , H4K20me1 increase on autosomes is approximately two-fold more compared to the increase on the X . Immunofluorescence studies in set-4 null mutants also showed an overall H4K20me1 increase resulting in the equalization of X and autosomal H4K20me1 levels [18 , 19] . Plotting H4K20me1 enrichment as measured by standard ChIP-seq analysis across transcription start sites illustrates the equalization of H4K20me1 levels between X and autosomes in dpy-21 ( e428 ) and set-4 ( n4600 ) mutants ( Figs 7D and S7A ) . In these plots , ChIP-seq experiments were performed without the spike-in control , and the genome-wide increase of H4K20me1 in the set-4 ( n4600 ) mutant was masked . Adjusting enrichment values by the spike-in ratios obtained from Fig 7C illustrates the inferred H4K20me1 levels in the set-4 ( n4600 ) mutant L3s ( lighter colors in Fig 7D ) . The ten-fold increase in H4K20me1 levels across the genome was surprising , based on the observation that western blot analysis of H4K20me1 showed approximately five-fold increase in the set-4 null mutant L3s ( Fig 4B ) . The basis of this difference between the two different assays ( ChIP versus western blot ) is unclear . A ten-fold increase of H4K20me1 may be possible in the light of previous studies in mammalian cells , which showed that ~80% of H4 is di-methylated at H4K20 [48 , 49] . If demethylation of H4K20me1 is inefficient in C . elegans , set-4 mutation may result in high levels of monomethylation . The magnitude of H4K20me1 increase and the difference between X and autosomes was greater in L3s compared to mixed stage embryos in set-4 ( n4600 ) mutant ( S7B Fig ) . This may be due to less H4K20me1 X-enrichment in embryos compared to L3s , consistent with H4K20me1 enrichment occurring later in embryonic development [18 , 29] . Incomplete X-enrichment of H4K20me1 in mixed stage embryos was also reflected by less H4K20me1 ChIP-seq enrichment on the X chromosome in wild type embryos compared to L3 ( Fig 7E , left panel ) . The delayed H4K20me1 enrichment may also cause a delay in full dosage compensation . X chromosome expression was not fully balanced between sexes during embryogenesis ( Fig 3A ) . While this may be due to differences in sex-biased expression between developmental stages ( see discussion ) , it may also be due to slightly less dosage compensation due to incomplete enrichment of H4K20me1 on the X chromosomes until the comma stage of embryogenesis . Indeed , X chromosome derepression upon dpy-27 depletion was slightly but significantly higher in L3 larvae compared to mixed stage embryos ( Fig 3E , Wilcoxon rank sum test p value = 5x10-7 ) . This was in spite of having more efficient DPY-27 knockdown in embryos compared to L3 larvae ( S4B Fig ) . X-enrichment of H4K20me1 may slightly strengthen DCC mediated X chromosome repression in later development . The small but significant X derepression observed in the set-4 ( n4600 ) mutant suggested that X/A balance of H4K20me1 contributes slightly to X chromosome dosage compensation ( Fig 4C ) . To address the mechanism of this contribution , we further analyzed gene expression in the set-4 ( n4600 ) mutant . Since mRNA-seq measures relative expression changes , we could not distinguish if the relative X derepression seen in the set-4 ( n4600 ) mutant is due to repression from autosomes or an increase from the X . Therefore , we adapted the spike-in normalization scheme for mRNA-seq analysis ( Fig 8A ) . To do this , we mixed 1 , 000 C . briggsae worms to 10 , 000 N2 ( wild type ) or set-4 mutant worms before RNA extraction and mRNA-seq analysis . We used fed L1 worms where we could reliably assume that the number of cells was the same between wild type and the mutant . The replicates correlated well with each other ( S8A Fig ) . Since spiked in C . briggsae worms were the same across all samples , we normalized C . elegans expression levels using the linear regression coefficients that equalized C . briggsae expression between wild type and mutant samples ( S8B Fig ) . Spike-in analysis of mRNA-seq levels in the set-4 ( n4600 ) mutant showed decreased expression from autosomes ( Fig 8B ) , resulting in the relative effect of X derepression seen in Fig 4C . To validate the spike-in mRNA-seq method , we compared it to RT-qPCR analysis for a number of genes . RT-qPCR was performed on the same total RNA preparations that were used for making the mRNA-seq libraries . Since RT-qPCR also relies on the assumption that total mRNA levels are the same between the wild type and the mutant , two C . briggsae genes were used as standard . Results from standard and spike in normalized RT-qPCR results were similar to mRNA-seq analysis ( Fig 8C ) . We reason that if H4K20me1 acts as a downstream effector of the DCC ( i . e . the effect of H4K20me1 on transcriptional repression was the same on the X and autosomes ) , equalization of H4K20me1 between the X and autosomes should result in the same level of X chromosome derepression . Two observations suggested that while H4K20me1 is important , X/A balance of H4K20me1 is not responsible for X chromosome repression . First , upon equalization of H4K20me1 levels between X and autosomes in the set-4 ( n4600 ) mutant , X chromosome is only slightly derepressed ( Fig 5C ) . Second , H4K20me1 depletion in the set-1 ( tm1821 ) mutant resulted in greater X derepression compared to equalization of H4K20me1 between X and autosomes in the set-4 ( n4600 ) mutant ( Fig 5C ) . Thus H4K20me1 does not act as the sole downstream effector of the DCC . Based on the spike in controlled ChIP-seq and mRNA-seq analyses in the set-4 ( n4600 ) mutant , we propose a model that H4K20me1 enrichment only slightly represses X chromosome transcription , but H4K20me1 may positively affect DCC activity , which is primarily responsible for X repression ( Fig 8E ) .
Our results indicate that in early embryogenesis , X chromosome is not fully dosage compensated in C . elegans . Then , how is the X chromosome dosage difference tolerated ? First , maternal loading of mRNAs into oocytes help equalize overall X-chromosomal transcript levels between the two sexes , thus genes that are important for the development of both sexes may be provided in the oocyte . Second , X chromosome harbors fewer genes with essential functions for embryogenesis [50] . Third , dosage sensitive genes such as haploinsufficient genes are depleted from the X chromosome [51 , 52] . We also found that C . elegans X chromosome has fewer genes that encode for multi-subunit protein complexes , which were found to be more dosage sensitive [53–55] . There are only 16 X-chromosomal genes with a yeast homolog of a multi-protein complex [52] , representing a significant depletion of such genes from the X ( p< 10−5 , Fisher’s exact test ) . For the remaining dosage sensitive genes on the X , in addition to DCC mediated repression , potential post-transcriptional mechanisms may act , as such regulation is extensive in C . elegans [56] . Thus , a combination of different mechanisms buffers the X chromosome dosage difference between hermaphrodite and males during early embryogenesis . Comparison of gene expression levels between hermaphrodite and male embryos reflects two genetic differences: X chromosome number and sex-biased gene expression . Others and we have shown that the X chromosome contains higher numbers of hermaphrodite-biased genes [51 , 57] . This may contribute to the higher X expression in hermaphrodites compared to the mixed sex embryos ( Fig 3 ) . However , most sexual dimorphism in C . elegans develops in the larval stages ( reviewed in [58] ) , and the adult gonads contribute an especially high amount of sex-biased gene expression [51] . Indeed , mRNA-seq replicates in the hermaphrodite and mixed-sex worms clustered by developmental stage rather than sex until the young adult stage ( S1A Fig ) . Thus , in embryos and larvae , developmental gene expression has a stronger effect on the transcriptome than sex biased gene expression . Nevertheless , there is sex-biased gene expression in embryos , including genes in the sex determination pathway ( Fig 2D ) . Interestingly , the XSEs were dosage compensated during late embryogenesis , suggesting that the DCC acts on these genes regardless of their early hermaphrodite-biased expression [36] ( S3 Fig ) . It is possible that the DCC mediated gene repression is chromosome-wide and superimposed onto other mechanisms of regulation that act at the level of individual genes , such as sex biased gene expression programs . Future studies that uncouple sex from X chromosome number are needed to resolve the relation between dosage compensation and sex-biased gene expression . Although X chromosome dosage compensation is essential for proper development , it is still not clear how DCC mutations cause lethality ( reviewed in [1] ) . In many animals , proper chromosome dosage is important , as trisomy or monosomy of chromosomes is not tolerated ( reviewed in [59] ) . X chromosome dosage compensation problems in mammals and worms would cause upregulation of X-chromosomal genes . In the case of trisomies , increased expression from a set of dosage sensitive genes may cause problems , as in Down syndrome ( reviewed in [60] ) . Additionally , increased expression from a whole chromosome pressures proteasome pathways needed to degrade the extra proteins , as was shown in yeast [61] . DCC mutations cause derepression of many X chromosomal genes that are involved in wide range of processes ( Fig 6 ) . Genes that are upregulated in the dpy-27 mutant include several transcription factors and kinases , whose overexpression may cause off-target effects ( discussed in [62] ) . While it is possible that the combination of a number of dosage sensitive genes on the X are responsible for the DCC phenotypes , DCC-mediated repression is applied across the X chromosome ( Fig 4 ) , suggesting that the DCC evolved to repress average X chromosomal gene dosage . DCC starts binding to the X chromosomes at around the 40-cell stage , which is followed by H4K20me1 enrichment on the X , which occurs between the 100-cell and comma stages [18 , 29] . Since the DCC reduces SET-4 activity , and SET-4 needs previous monomethylation of H4K20 to catalyze di- and trimethylation , H4K20me1 enrichment on the X can only happen after the initial deposition of H4K20me1 during cell division . The delay in H4K20me1 enrichment during embryogenesis may be due to a requirement for cell divisions to initially deposit H4K20me1 . In mammalian tissue culture cells , cell-cycle regulation of SET-1 ensures a dramatic increase in H4K20me1 on mitotic chromosomes [63 , 64] . H4K20me1 also increases on mitotic chromosomes in C . elegans [18 , 29] . Therefore , it is possible that multiple divisions are necessary to successively increase H4K20me1 on the C . elegans X chromosomes . A microscopy-based study of H4K20me1 enrichment across multiple divisions in one cell lineage can address this possibility . Previously , we hypothesized that the DCC increases H4K20me1 by reducing SET-4 activity on the X chromosomes [18] . Analysis of gene expression in the set-4 null mutant suggested an effect on X chromosome dosage compensation ( Fig 5C ) . This effect was indirect , as an average of ~20% reduction across the autosomes resulted in a relative X chromosome derepression compared to autosomes ( Fig 8B ) . Since H4K20me1 increase in the set-4 null mutant resulted in repression , H4K20me1 enrichment on the X chromosome may slightly contribute to repression in hermaphrodites . H4K20 methylation has a role in gene expression in other organisms as well , but the mechanisms remain unclear [65] . It is possible that H4K20me1 positively regulates chromatin compaction , thus reduces accessibility of RNA Polymerase II to the X chromosomes . H4K20me1 is enriched on mitotic chromosomes and the inactive X chromosome in mammals [65] . In C . elegans , H4K20me1 is enriched on mitotic chromosomes and across the entire X ( Fig 7A ) , which is more compact in hermaphrodites [66] . Both set-1 and set-4 resulted in a relative increase in X chromosome volume compared to the autosomes in hermaphrodites [66] . While set-4 is important for X compaction , the null mutant showed only a slight effect on X chromosome dosage compensation and does not have obvious dosage compensation phenotypes in standard laboratory conditions ( Fig 5C ) . Therefore , X compaction by itself does not fully explain DCC mediated X chromosome repression . Regulation of H4K20 methylation by set-4 may be involved in additional processes . Set-4 knockdown increased lifespan in an RNAi screen [67] and suppressed the developmental delay associated with mutant rict-1 , a component of the target of rapamycin complex 2 [68] . Gene expression changes that occur in the set-4 mutant may be the basis for these phenotypes . Alternatively , set-4 may act through other non-histone substrates , as recent experiments in D . melanogaster showed that H4K20A mutation did not completely recapitulate the phenotype of the methyltransferase mutant [69] . Our working model for H4K20me1 in X chromosome dosage compensation is summarized in Fig 8E . Briefly , X-enrichment of H4K20me1 contributes slightly to X chromosome repression , but H4K20me1 by itself is not a major downstream effector of the DCC . Instead , H4K20me1 level is important for DCC-mediated repression of the X chromosomes . One possibility is that H4K20me1 may increase association of condensin DC with chromatin . In agreement with this possibility , there was slightly less DPY-27 binding in the dpy-21 ( e428 ) mutant embryos where H4K20me1 is no longer enriched on the X chromosome ( S7C Fig ) . In mammalian cells , HEAT domains found in N-CAPD3 and N-CAPG2 subunits of condensin II bind to H4K20me1 containing histone tail peptides , and this interaction was proposed to increase condensin II binding on mitotic chromosomes , which are enriched for H4K20me1 [70] . The DCC contains two subunits with HEAT domains , DPY-28 and CAPG-1 [20] . Positive regulation of DCC by H4K20me1 may help maintain dosage compensation complex on the X in later developmental stages in C . elegans . Such maintenance mechanisms are involved in epigenetic gene regulation , and have been demonstrated to be important for mammalian X inactivation [71 , 72] . Previous work and our results suggest that the DCC represses X chromosome transcription . The molecular mechanism by which the DCC represses transcription remains unclear . In various organisms , condensins demonstrate activities such as positive supercoiling [73–75] , single stranded DNA reannealing [76 , 77] and making topological links in DNA [78 , 79] . It is not known if condensin DC has these activities . The DCC is enriched at active promoters [80] and the DCC reduces RNA Pol II binding to X chromosome promoters [43] . The DCC ChIP binding scores at promoters positively correlate with the amount of transcription [81] . Positive correlation between condensin binding and transcription was also seen in yeast and mammalian cells [77 , 82–84] . We had hypothesized that coupling the level of DCC binding to the level of transcription at each promoter could be the basis of the two-fold reduction in gene expression regardless of the expression level [80] . Indeed , GRO-seq analysis indicated that while the DCC binding positively correlates with transcription , the level of DCC mediated reduction in RNA Pol II is similar across genes transcribed at a wide range of levels [43] . Based on the observation that the DCC binds to promoters , and that yeast condensin interacts with TBP [83] , it is possible that the DCC is recruited to promoters by a general transcription factor , and then halves the rate of RNA Pol II binding at each cycle of transcription . Alternative but not exclusive DCC mechanisms have been proposed , such as compacting chromatin to reduce activators’ access to the X chromosome or changing interactions between enhancer and promoters to reduce RNA Pol II recruitment [43] . These models are based on the observations that condensins are required for chromosome condensation , and in other organisms such as D . melanogaster and mammalian cells , condensins regulate long-range chromosomal interactions [85 , 86] . Recently , the DCC was shown to compact the X chromosome [27] , regulate sub-nuclear localization of the X [87] , and form topologically associated domains on the X [88] . However , it remains unclear if these large-scale chromatin changes are the cause or consequence of increased transcription from the X chromosomes in the DCC mutants . In mouse , C . elegans and D . melanogaster , dosage compensation complexes are targeted to and regulate the chromatin structure and transcription from the X chromosomes in one of the two sexes ( reviewed in [1] ) . The composition of the dosage compensation complexes suggests that different complexes coopted different sets of gene regulatory mechanisms to the X chromosomes . In the case of C . elegans , duplication and divergence of a condensin subunit ( dpy-27 ) created the dosage compensation specific condensin complex ( condensin DC ) that reduces transcription from both X chromosomes in hermaphrodites . Condensins are essential for chromosome condensation during cell division , and regulate chromatin compaction and transcription during interphase . H4K20me1 is a mitosis enriched chromatin mark that is also enriched on the dosage compensated X chromosomes in C . elegans . Our work on the DCC and H4K20me1 contributes to understanding the link between the mitotic and interphase function of condensins in chromosome condensation and gene regulation .
N2 , wild type; BS553 , fog-2 ( oz40 ) ; CB428 , dpy-21 ( e428 ) V; MK4 dpy-27 ( y56 ) /qC1[dpy-19 ( e1259 ) glp-1 ( q339 ) ] [qIs26] III; SS1075 set-1 ( tm1821 ) /hT2g[bli-4 ( e937 ) let- ? ( q782 ) qIs48] ( I;III ) ; MT14911 set-4 ( n4600 ) II; VC199 , AF16 C . briggsae wild isolate . Unless otherwise noted , strains were maintained at 20°C on NGM agar plates using standard C . elegans growth methods . For staged embryo collections , worms were synchronized by bleaching gravid worms , allowing embryos to hatch overnight and growing the synchronized L1s on NGM plates for ~60 hours until the first embryos were seen in hermaphrodite gonads . Early embryos were isolated from bleached young adults , and immediately collected . Comma embryos were obtained by aging early embryos 4 hours in M9 buffer . Larvae were collected after growing hatched L1s for 6 hours ( L1 larvae ) or 24 hours ( L3 larvae ) on NGM plates . Young adults were synchronized by growing hatched L1s for 44 hours at 25°C , and collected prior to the accumulation of fertilized embryos in the gonad . Mixed stage embryos were collected by bleaching gravid adults . Developmental stages were assessed by counting DAPI stained nuclei by DAPI staining , assessing germline , male tail and vulva morphology . Due to lethality , set-1 ( tm1821 ) mutants were kept heterozygous with a balancer chromosome that includes a GFP marker ( SS1075 strain [18] ) . set-1/balancer heterozygous larvae were selected by picking GFP ( + ) animals and set-1 homozygous mutant larvae were identified by picking GFP ( - ) animals . Since dpy-27 ( y56 ) is lethal [12 , 28] , dpy-27 ( y56 ) worms were kept as heterozygotes with a balancer chromosome that encodes rol-6 ( su1006 ) . dpy-27 ( y56 ) homozygote embryos were collected by bleaching young adults without the roller phenotype . This resulted in embryos that lacked both maternal and zygotic dpy-27 . For DPY-27 RNAi , the bacterial strain from the Ahringer RNAi library containing dpy-27 RNAi plasmid and control plasmid were grown and induced in 100 ml LB media for 3 hours with 0 . 1 mM ITPG , concentrated 10-fold and seeded onto a 10 cm NGM plate supplemented with ampicillin , tetracycline and 1mM IPTG . Synchronized L1s were placed on seeded plates and collected at 24 hours ( L3 larvae ) or grown to gravid adults and embryos were collected by bleaching ( mixed stage embryos ) . At collection time , all samples were washed 3 times in M9 buffer and stored in ten volumes of Trizol ( Invitrogen ) at -80°C . 2–3 collections were pooled for one biological replicate . A summary of all the data sets and replicates are provided in S1 File . Raw data files and wiggle tracks of ChIP enrichment per base pair , and RNA-seq FPKM values are provided at Gene Expression Omnibus database ( http://www . ncbi . nlm . nih . gov/geo/ ) under accession number [GSE67650] . For RNA preparation , samples in Trizol were frozen and thawed 3–5 times to freeze-crack worms , and phase containing the total RNA was cleaned up using the Qiagen RNeasy kit . mRNA was purified using Sera-Mag oligo ( dT ) beads ( ThermoScientific ) from 0 . 5–10 ug of total RNA . Stranded mRNA-seq libraries were prepared based on incorporation of dUTPs during cDNA synthesis using a previously described protocol [89] . Single-end 50-bp sequencing was performed using the Illumina HiSeq-2000 . Reads were mapped to the C . elegans genome version WS220 with Tophat v1 . 4 . 1 [90] using default parameters . Gene expression was quantified using Cufflinks version 2 . 0 . 2 [91] for strand-specific reads using default parameters and supplying gene annotations . Gene expression values used were the mean of at least three biological replicates , and are provided in S2 File . Differential expression analysis was performed using DESeq2 version 1 . 0 . 17 [92] in R version 3 . 0 . 2 , and the values are given in S3 File . All analyses were performed with genes that had average expression level above 1 FPKM ( fragments per kilobase per million , as calculated by Cufflinks ) . RNA-seq data clustering was performed by K-means clustering using the Hartigan-Wong method in R version 3 . 02 . GO term analysis was performed using GOrilla [93] . The clusters and GO analyses results are provided in S4 File . ChIP-seq was performed as previously described [26] , using H4K20me1 antibody ( Abcam ab9051 ) , a custom DPY-27 antibody [80] , and AMA-1 antibody against the large subunit of RNA-Pol II ( modENCODE SDQ2357 ) . Single end reads were aligned to the C . elegans genome version WS220 using Bowtie 1 . 0 . 1 [94] , allowing up to two mismatches in the seed , returning only the best alignment , and restricting a read to map to at most four locations in the genome . Peak calling and genome-wide coverage estimation was obtained with MACS version 1 . 4 . 2 [95] using mapped reads from ChIP and input . To estimate the final genome coverage , coverage per base was normalized to the genome-wide median coverage , excluding the mitochondrial chromosome , and the input was subtracted from the ChIP . Replicates were merged by averaging coverage at each base position across replicates . For peak calling , ChIP and input reads were merged from all replicates using samtools merge version 0 . 1 . 19 [96] , and MACS was used for peak calling . The peak calling was also done per replicate , and peaks from the combined data that also occur in the majority of replicates were retained . For wild type and mutant comparisons , data sets were standardized by z-score transformation of the ChIP enrichment values based on the presumed background , excluding regions covering MACS peaks ( e-5 cutoff ) . The statistical framework for identifying differentially expressed genes takes into account the average and standard deviation between replicates . Genes that are not found as differentially expressed cannot be labeled as “statistically” similarly expressed , because variability between replicates may cause these genes to fail the statistical test . Therefore , to determine genes that are similarly expressed in hermaphrodite and mixed sex embryos , we constructed a 95% confidence interval around the log2 expression ratio for each gene . For a gene to be called similarly expressed , that confidence interval could not exceed a 30% change in expression . Distribution of similarly , equivalently and differentially expressed results are provided in S5 File . To determine if there is two classes of DCC regulated genes on the X chromosome , we used an expectation-maximization algorithm to determine the distribution of expression changes upon dpy-27 knockdown . A mixture distribution with two or three components was assumed and the better fit was used . Parameters of the mixture distribution were estimated using the mixtools package [97] in R version 3 . 0 . 2 . For RNA-seq , hatched L1s were fed on OP50 seeded plates for 6 hours , collected , counted three times and diluted to the desired number of worms based on the average count . 10 , 000 C . elegans L1s were mixed in with 1 , 000 C . briggsae L1s . RNA-seq was performed as described above . mRNA-seq reads were mapped to the C . elegans genome version WS220 and C . briggsae genome version WS224 using TopHat ( Trapnell et al 2010 ) version 2 . 0 . 11 and reads mapping to both genomes were discarded from further analyses . Raw counts for each gene were estimated with htseq-count version 0 . 6 . 1 [98] . All read counts were normalized based on total reads and the median across three replicates was determined . A linear model was fitted to the highly expressed genes from wild type and mutant C . briggsae spike-in data , whose coefficients were used to correct the C . elegans expression data . For ChIP-seq , 0 . 1–0 . 2 mg of ChIP-extract prepared from C . briggsae embryos were added to 1–2 mg C . elegans ChIP extract . ChIP and corresponding input reads were mapped to the C . elegans genome version WS220 and C . briggsae genome version WS224 using Bowtie version 1 . 0 . 1 [94] and reads mapping to both genomes were discarded . To determine a spike-in normalization factor , mapped reads were first normalized to total reads and then the ratio between ChIP and input was calculated for uniquely mapped reads to C . elegans or C . briggsae . The ratio of C . elegans and C . briggsae ChIP/Input between the mutant and wild type was used to correct the genome coverage in the mutant . We used the total RNA that was isolated for the spike in libraries . 20 nanograms of total RNA was used per 20 microliter one-step RT-qPCR reaction ( KAPA Biosystems ) . To eliminate cross amplification between C . elegans and the C . briggsae spike in RNA , primers were designed to amplify C . elegans that do not amplify in the C . briggsae genome . C . briggsae genes CBG27711 and CBG25839 were selected as the spike in normalization gene based on L1 expression , lack of C . elegans homolog and lack of amplification from the C . elegans template cDNA . No-RT and opposite species template RNA reactions were used as controls to confirm specificity of primers . Relative fold change between set-4 ( n4600 ) and wild type was calculated using the ΔCt method and spike in normalized changes were calculated using the ΔΔCt method using the averageΔCt of C . briggsae genes CBG27711 and CBG25839 between wild type and set-4 as the second normalization parameter [99] . Primer sequences are as follows: kin-3 forward TTTCAAGGGACCCGAGCTTC , kin-3 reverse GTGTCGCCCGAGAATATCGT , pmp-2 forward CCTGGAGTGGTTGGAATGCT , pmp-2 reverse TCTTGCCAGTTTGCCCAGAA , hlh-11 forward TTCGTTCGGATAGCGCAGAA , hlh-11 reverse GGTGCCATTCGTGCATTTGT , tag-115 forward GTTCAGGACCCCGGTCAAAA , tag-115 reverse TTGGAAGTGATCGCTGACCG , rig-3 forward GCACCAGGAAACAAGCAAGG , rig-3 reverse TTGCATCACGGACGTGGTTA , ifp-1 forward TCCAGAGCGACAGTAGTGGA , ifp-1 reverse TAGTAGTTCGTGCCGGCTTC , rgs-6 forward AATGTGAGCGAACCCAGCAA , rgs-6 reverse CGCCTCGTCGCCTATTTTTG , dhs-25 forward CTGGGAGGACGAATTGCTGT , dhs-25 reverse CCCCTTCACACTGTCTGCAT , CBG27711 forward ATGAGCATAATTCAGAGAGAAAATTTT , CBG27711 reverse CTAGGCATCTATAATTAGGTT , tag-261 forward CGTAGATGCACGTCTTCAAGC , tag-261 reverse TCTCCGCCTCTTTGAGACCTC , C12D5 . 09 forward TGACTTTTTCTCGACGGATTTG , C12D5 . 09 reverse CACACCTTCTTTCGTTGGCG , C06C3 . 12 forward TCGGTGGACCGATTCTAATGAA , C06C3 . 12 reverse TTAGCCCACTTCGACTGAGG , B0507 . 10 forward TCTCAGGAATTGCTAATTCGGC , B0507 . 10 reverse GGCTCTGTGGCGGTTGATAA , tatn-1 forward CGCTTCTTGAGCAAGCCAAA , tatn-1 reverse CTGCGATCTCACTTGGTGGA , nspc-19 forward CATCCTTGCTGCTCTCTGCT , nspc-19 reverse TTCTCCACCGTTGACACCTG , B0294 . 1 forward ATCCGCAACCAACTCTCCTG , B0294 . 1 reverse CAAGGTTTTCTCCATTCTTTGGCA , CBG25839 forward CGCGTCTAAACCAGCCAAAC , CBG25839 reverse AGCTGAGCGAGATGAACACC . | In many animals , males have a single X and females have two X chromosomes , creating an X chromosomal gene dosage imbalance between sexes . This imbalance is corrected by dosage compensation mechanisms , which are essential for development in mammals , flies and worms . The timing and the molecular mechanisms of dosage compensation during development remain unclear . In Caenorhabditis elegans , a complex of proteins called the dosage compensation complex ( DCC ) binds to and represses transcription of both X chromosomes in hermaphrodites ( XX ) by half to match X expression to that of males ( XO ) . We found that the DCC starts repressing X chromosomes in early embryogenesis , but average X-linked transcript levels remain higher in hermaphrodite embryos compared to males until the larval stage . Later in embryogenesis , the DCC increases H4K20 monomethylation on the X chromosomes , which is important for dosage compensation . Our results suggest that H4K20 monomethylation does not carry out the transcriptional repressive function by itself but strengthens the DCC’s effect . In summary , the DCC localization starts X chromosome repression in early embryogenesis , and in later embryogenesis sets up a chromatin environment that aids dosage compensation that continues through development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Developmental Dynamics of X-Chromosome Dosage Compensation by the DCC and H4K20me1 in C. elegans |
Co-evolution of transcriptional regulatory proteins and their sites of action has been often hypothesized but rarely demonstrated . Here we provide experimental evidence of such co-evolution in yeast silent chromatin , a finding that emerged from studies of hybrids formed between two closely related Saccharomyces species . A unidirectional silencing incompatibility between S . cerevisiae and S . bayanus led to a key discovery: asymmetrical complementation of divergent orthologs of the silent chromatin component Sir4 . In S . cerevisiae/S . bayanus interspecies hybrids , ChIP-Seq analysis revealed a restriction against S . cerevisiae Sir4 associating with most S . bayanus silenced regions; in contrast , S . bayanus Sir4 associated with S . cerevisiae silenced loci to an even greater degree than did S . cerevisiae's own Sir4 . Functional changes in silencer sequences paralleled changes in Sir4 sequence and a reduction in Sir1 family members in S . cerevisiae . Critically , species-specific silencing of the S . bayanus HMR locus could be reconstituted in S . cerevisiae by co-transfer of the S . bayanus Sir4 and Kos3 ( the ancestral relative of Sir1 ) proteins . As Sir1/Kos3 and Sir4 bind conserved silencer-binding proteins , but not specific DNA sequences , these rapidly evolving proteins served to interpret differences in the two species' silencers presumably involving emergent features created by the regulatory proteins that bind sequences within silencers . The results presented here , and in particular the high resolution ChIP-Seq localization of the Sir4 protein , provided unanticipated insights into the mechanism of silent chromatin assembly in yeast .
Among all specialized chromatin structures , the difference between heterochromatin and euchromatin is perhaps the most fundamental , motivating intense study of the differences between these two structures . DNA sequences within heterochromatic regions evolve rapidly in animals [1] , [2] , plants [3] , [4] , and fungi [5] , presenting a paradox of how the specification of heterochromatin structure persists despite rapid changes in the underlying sequence [6] . In Saccharomyces the biology of heterochromatin has proven eminently accessible to genetic studies through its role in gene silencing [7] , and comparative studies of silencing now seem poised to illuminate key processes underlying heterochromatin evolution . Molecular co-evolution of transcriptional regulatory proteins with their sites of action has been proposed to maintain regulatory functions across species divergence [8] , [9] . In this context , “co-evolution” is typically understood as compensatory changes in a DNA sequence motif and the DNA-binding domain of the cognate transcription factor . Although it has been suggested that such co-evolution is prevalent in nature [8] , in only a few instances has it been directly tested [10]–[12] . In Dipteran insects , for example , co-evolution of bicoid binding sites in the hunchback promoter and the bicoid homeodomain has been proposed to maintain hunchback-mediated developmental patterning along the anterior/posterior axis in Musca and Drosophila [13] , [14] . However , the large size and complexity of animal regulatory elements , and the difficulty of performing cross-species complementation tests in animals , have precluded clear distinction between regulatory divergence and bona fide co-evolution . Transcriptional silencing by Sir ( Silent Information Regulator ) proteins is necessary for the specialized haploid mating-type system found in Saccharomyces [15] , [16] . DNA regulatory elements termed “silencers” contain binding sites for the Origin Recognition Complex ( ORC ) , Rap1 , and Abf1 , which in turn direct the assembly of silent chromatin structures at the HML and HMR loci . The current model for the establishment of silencing holds that a Sir2/Sir3/Sir4 complex is brought to silencers by protein-protein interactions between ORC and Sir1 , and between Rap1 and Sir4 [7] . Upon nucleation of these complexes , silent chromatin formation is catalyzed by the histone deacetylase activity of Sir2 , and propagated , at least in part , through interactions between Sir3 and newly deacetylated histone tails [17]–[19] . Sir proteins are not thought to bind specific DNA sites; instead , efficient nucleation of silencing complexes at silencers requires interactions between Sir1 and Sir4 , bridging the ORC-Sir1 and Rap1-Sir4 interactions [20] . Silencing also occurs at telomeres , which recruit Sir proteins primarily through arrays of Rap1 binding sites within the terminal repeats ( TG1–3 ) [21] . However , more complex regulatory elements reside near the terminal repeats , and at some telomeres these may also serve to recruit Sir proteins [32] , [23] . We have recently shown that silencer elements are among the most rapidly evolving regulatory sequences in Saccharomyces genomes [5]; however , the regulatory proteins that directly bind silencers are highly conserved , essential proteins . Intriguingly , the Sir1 and Sir4 proteins parallel the silencers in their rapid evolution , but these proteins show distinct patterns of evolution . SIR1-related genes have undergone multiple duplication and loss events: for example , S . bayanus has four functional paralogs of the single S . cerevisiae SIR1 gene , including the ancestral KOS3 ( Kin Of Sir1 ) paralog , which S . cerevisiae has lost along with two other paralogs ( Figure 1A ) [24] . In contrast , the Sir4 protein is among the 40 most diverged proteins between S . cerevisiae and S . bayanus ( Figure 1B ) , with 45% identity between its orthologs relative to a genome-wide average of 83% identity [25] , [26] . Although SIR2 and SIR3 each have a paralog resulting from the whole-genome duplication ( SIR2 has three additional , more ancient paralogs ) , neither gene has experienced subsequent duplication or loss events [27] , [28] . S . cerevisiae and S . bayanus are post-zygotically isolated—haploids of these two species can mate to form mitotically stable hybrid diploids , but meiotic spores derived from these diploids are usually inviable [29] , [30] . The rapid evolution of the silencers , the Sir4 protein sequence , and the elaboration of Sir1 paralogs make these two species an excellent phylogenetic context for comparative studies of silencing . Here , we describe functional studies in S . cerevisiae/S . bayanus interspecies hybrids that demonstrated how co-evolution among two heterochromatin proteins , Sir1 and Sir4 , and multiple silencer DNA elements allowed two divergent lineages to maintain robust silencing despite these rapid genetic changes . This example of regulatory co-evolution is of particular interest because the co-evolving proteins are not the agents that directly bind to the divergent regulatory DNA sites .
In the course of a genetic screen for S . bayanus silencing mutants , we discovered that S . cerevisiae SIR4 failed to complement S . bayanus sir4Δ mutants for silencing of both HML and HMR , but S . cerevisiae SIR2 and SIR3 complemented mutations in S . bayanus orthologs ( Figure S1; Zill et al . in preparation ) . This result was unanticipated as there are many cases of human proteins that can replace their yeast counterparts , even for proteins that function in large complexes and have considerably more sequence divergence than that seen between S . cerevisiae and S . bayanus proteins [31]–[34] . The incompatibility was unidirectional as S . bayanus SIR2 , SIR3 , and SIR4 complemented S . cerevisiae sir2Δ , sir3Δ , and sir4Δ , respectively . Importantly , SIR4 functional divergence was due to one or more coding changes , as the level of expression of the two Sir4 orthologs , measured at either the RNA or protein level , was equivalent ( Figure S2 ) . To assay the function of both species' silencing machineries in the same cellular milieu , we developed a highly sensitive transcriptional reporter assay in S . cerevisiae/S . bayanus interspecies hybrid diploids that allowed us to monitor silencing of each species' HMR locus ( hereafter referred to as Sc-HMR or Sb-HMR ) . The reporter consisted of the K . lactis URA3 open reading frame placed under the control of the endogenous HMRa1 promoter of each species , in two separate , but otherwise isogenic , hybrid strains ( Figure 2A ) . In these hybrids the S . cerevisiae SIR4 ( Sc-SIR4 ) allele could not , on its own , silence Sb-HMR ( Figure 2A , row 2 ) . Reduced dosage of Sir4 per se did not cause loss of silencing at Sb-HMR , as S . bayanus diploids with only one copy of Sb-SIR4 showed no silencing defect ( Figure 2B , row 2 ) , nor did S . cerevisiae diploids with only one copy of Sc-SIR4 ( unpublished data ) . Furthermore , a hybrid diploid containing two copies of Sc-SIR4 ( the Sb-SIR4 gene was replaced by Sc-SIR4 ) also failed to silence Sb-HMR ( Figure 2A , row 5 ) . In contrast , one Sc-SIR4 gene was able to silence Sc-HMR in all hybrid strains tested ( Figure 2A , bottom panel; Figure 3A ) . Thus , the hybrid cellular environment did not interfere with Sc-Sir4 function , and within a species , SIR4 was not haplo-insufficient . It appeared that Sc-Sir4 was either inhibited at Sb-HMR by something encoded by the S . bayanus genome specifically or somehow failed to interact with proteins that promoted Sb-HMR silencing . Transcription analysis of a critical set of the hybrid strains showed good correspondence between expression of the HMR::URA3 reporter and growth patterns observed on FOA and CSM/-Ura media ( Figure 3B ) . We note that in the interspecies hybrids with both species' SIR4 alleles ( Sc-SIR4/Sb-SIR4 ) , Sb-HMR silencing appeared weakly defective relative to the complete silencing of Sb-HMR in S . bayanus diploids by both the reporter assay and direct RNA measurement ( compare Figure 2A , row 1 with Figure 2B , row 1; Figure 3B ) . In addition , Sb-HMR silencing was further weakened in hybrids lacking Sc-Sir4 ( Figure 2A , compare row 3 with row 1 ) . This result was paradoxical because Sc-Sir4 appeared to have very little ability to silence Sb-HMR in hybrids lacking Sb-Sir4 . As explained below , these weak Sb-HMR silencing defects were likely due to an emergent property of the hybrids , resulting from unusually strong interactions between Sb-Sir4 and S . cerevisiae silent loci that effectively reduced Sb-Sir4 associations with Sb-HMR . The presence of Sc-Sir4 limited the competition for Sb-Sir4 . The inability of Sc-Sir4 to function at Sb-HML and Sb-HMR could have been manifested either during its recruitment or after its assembly into chromatin [35] . To determine where in the assembly of S . bayanus silenced chromatin Sc-Sir4 protein was blocked , we compared the ability of Sc-Sir4 and Sb-Sir4 proteins to associate with all silent loci of both species at high resolution using chromatin-immunoprecipitation followed by deep-sequencing of the precipitate ( ChIP-Seq ) . Sir4 ChIP-Seq was performed using hybrid diploids expressing Sc-Sir4 only , Sb-Sir4 only , or both Sc-Sir4 and Sb-Sir4 . Because of the sequence divergence between HML and HMR of the two species , the occupancy of each species' HML and HMR loci could be evaluated simultaneously . In each strain , only one SIR4 allele carried a 13xMyc epitope tag [36] . In hybrids expressing Sc-Sir4 only , robust enrichment of Sc-HML and Sc-HMR silencers was observed as expected , with very weak enrichment of Sb-HML and Sb-HMR silencers ( Figure 4A , Table 1 ) . Strikingly , Sc-Sir4 association with an internal region of Sb-HMR was indistinguishable from non-silenced regions . In contrast , as predicted from the genetic results , Sb-Sir4 associated robustly with HML and HMR loci from both species , and did so most robustly at S . cerevisiae silencers ( Figure 4A , Table 1 ) . The ChIP-Seq results were validated at Sc-HMR , Sb-HMR , and control loci using standard ChIP-qPCR analysis ( Figure 5A ) . Thus , Sc-Sir4 showed strongly reduced association with Sb-HML and Sb-HMR silencers and no detectable association with their internal regions . The relative absence of Sc-Sir4 from these normally silenced regions of the S . bayanus genome was consistent with two possibilities . Perhaps Sc-Sir4 could not interact properly with Rap1 , ORC , or the S . bayanus Sir1 paralogs assembling on their silencers , or perhaps an S . bayanus protein was preventing stable association between Sc-Sir4 and S . bayanus silencers . The comparative Sir4 ChIP-Seq data provided a surprising insight into the mechanism of Sir4 incorporation into silent chromatin . Although Sc-Sir4 binding to Sb-HML and Sb-HMR loci was barely detectable in hybrids expressing Sc-Sir4 only ( Figure 4A , center and right panels; Table 1 ) , in hybrids expressing both Sc-Sir4 and Sb-Sir4 , Sc-Sir4 binding increased substantially at Sb-HML and Sb-HMR silencers and internal regions ( Figure 4A; Table 1 ) . Thus , despite the poor ability of Sc-Sir4 to associate with Sb-HML and Sb-HMR on its own , Sb-Sir4 somehow provided Sc-Sir4 access to them . It appeared that Sir4 association with S . bayanus HML and HMR involved two distinguishable modes of interaction , but Sc-Sir4 was capable of only one ( Sb-Sir4-dependent ) . Moreover , the divergent mode was apparently critical only for the initial association of Sir4 with a silencer , and not for subsequent associations with the silenced region . The ability of Sc-Sir4 to form dimers suggested one straightforward explanation for the Sb-Sir4-dependent chromatin association: inter-specific Sc-Sir4/Sb-Sir4 dimerization through a conserved coiled-coil domain [37] , [38] . Sb-Sir4-assisted incorporation of Sc-Sir4 into Sb-HML and Sb-HMR was consistent with Sc-Sir4 contributing to silencing at these loci , as suggested by the decreased Sb-HMR silencing in hybrids lacking Sc-Sir4 ( Figure 2A , row 3 ) . However , this hypothesis per se could not explain the sensitivity of Sb-HMR silencing to reduced Sb-SIR4 dosage that was observed in interspecies hybrids , but not in S . bayanus diploids ( compare Figure 2A , rows 1 and 3 , with Figure 2B , row 2; Figure 3B ) . Further analysis of Sir4 localization on the S . cerevisiae/S . bayanus hybrid genome by ChIP-Seq provided an explanation of this hybrid-specific Sb-SIR4 dosage sensitivity , as described next . Given the differential association of Sc-Sir4 and Sb-Sir4 with the two species' HML and HMR loci , we asked if any other loci , genome-wide , also showed a dramatic discrepancy . In S . cerevisiae , silencing by Sir proteins occurs at telomeres and subtelomeres , in addition to HML and HMR [18] , [39] , [40] . A comparison of the interspecies hybrids expressing Sc-Sir4 only versus Sb-Sir4 only showed that all S . cerevisiae TG1–3 terminal repeats ( which contain embedded Rap1 binding sites ) , including those present on the centromere-proximal side of some Y′ elements , were comparably occupied by both species' Sir4 proteins ( Figure 4B ) . ( Y′ elements are helicase-encoding repetitive sequences of unknown origin and function that are found in some subtelomeric regions immediately adjacent to the terminal repeats [22] . ) This result was not surprising as the telomerase-replicated repeated sequence , templated by the TLC1 RNA , is identical in the two species ( our unpublished observations ) . Thus , it appeared that Sir4 association with the S . cerevisiae genome , as promoted by Rap1 , was not substantially different between Sc-Sir4 and Sb-Sir4 . Indeed , the C-terminal residues of Sc-Sir4 critical for its interaction with Rap1 are conserved in Sb-Sir4 ( our unpublished observations ) . We note that the smaller ChIP-Seq peaks observed in these regions in the “No tag” control strain ( Figure 4B , yellow shading ) are likely due to non-specific DNA binding to the anti-myc beads . Unexpectedly , S . cerevisiae subtelomeres had two types of regions notably more enriched by Sb-Sir4 ChIP than by Sc-Sir4 ChIP . These regions corresponded to X elements , which are regulatory sequences near telomere ends that contain ORC and Abf1 binding sites [22] , and the ORFs within Y′ elements . For X elements , ChIP-Seq of Sc-Sir4 showed an average of 7-fold enrichment , whereas Sb-Sir4 showed an average of 14-fold enrichment , with even greater disparity often evident immediately adjacent to X elements ( Figure 4B ) . Therefore , Sb-Sir4 either associated more robustly with factors bound to X elements than did Sc-Sir4 or conceivably was excluded less effectively . X element core sequences ( containing the ORC and Abf1 binding sites ) are bordered on the telomere-proximal side by X-element combinatorial repeats ( formerly known as subtelomeric repeats or STRs; [22] ) and the terminal repeats ( see http://www . yeastgenome . org/images/yeastendsfigure . html for schematics depicting X-only and X-Y′ telomere ends ) . The differential pattern of Sir4 association with X elements was consistent with Sb-Sir4 associating more robustly than Sc-Sir4 with sequences at , and immediately adjacent to , the ORC binding sites , presumably via ORC-mediated interactions ( Figure 4B ) . Other S . bayanus proteins produced in the hybrids , such as the Sir1 paralogs , may contribute to the enhanced association of Sb-Sir4 with X elements , as discussed below . We observed weak Sc-Sir4 association with Y′ elements despite its strong association with neighboring terminal repeats ( Figure 4B , right panel ) , consistent with earlier observations using ChIP-chip and transcription reporter analyses [23] , [41] . Surprisingly , Sb-Sir4 associated considerably better than Sc-Sir4 with all Y′ elements , showing an average of 5-fold enrichment across their coding regions by Sb-Sir4 ChIP versus 1 . 2-fold enrichment by Sc-Sir4 ChIP . We note that the S . bayanus genome lacks Y′ elements , and thus S . bayanus subtelomeres may have reduced Sir4 recruitment potential relative to S . cerevisiae subtelomeres [42] , [43] . Thus , the enhanced associations of the Sb-Sir4 protein with X and Y′ elements suggested that , in the hybrid strains , S . cerevisiae telomeres might have competed with Sb-Sir4 association with Sb-HML and Sb-HMR , leading to the somewhat weakened Sb-HMR silencing observed in hybrids with only one copy of Sb-Sir4 ( Figure 2A , rows 1 and 3; Figure 3B ) . Sb-Sir4 association was indeed reduced at Sb-HMR and Sb-HML silencers in a hybrid expressing only one copy of Sb-SIR4 , relative to a hybrid with two copies of Sb-SIR4 ( Table 1 , compare columns 2 and 4 ) . Thus , Sc-Sir4 may have , in effect , protected Sb-HMR silencing in hybrids when Sb-Sir4 was present ( Figure 2A , compare rows 1 and 3 ) by occupying sites at S . cerevisiae telomeres that would otherwise have been bound by Sb-Sir4 . ( Although the S . cerevisiae Y′ elements are bound by Sb-Sir4 and not by Sc-Sir4 in cells with only a single species' Sir4 , in the Sc-SIR4/Sb-SIR4 hybrids , Sc-Sir4 and Sb-Sir4 both occupy Y′ elements ( unpublished data ) . However , the extent of occupancy by Sb-Sir4 is less than in cells with Sb-Sir4 only , consistent with Sc-Sir4's ability to spare Sb-Sir4 binding to Y′ elements in the hybrids . ) The ChIP-Seq data allowed us to determine whether the species restriction to Sc-Sir4 association , evident at Sb-HML and HMR , also applied to S . bayanus telomeres . Although subtelomeric regions of the S . bayanus genome are presently incompletely assembled and annotated ( see Saccharomyces Genome Database , www . yeastgenome . org ) , we identified several candidate subtelomeric contigs based on homology to S . cerevisiae subtelomeric genes and X elements . Contigs from the S . bayanus genome assembly that contained regions bound by both Sc-Sir4 and Sb-Sir4 ( as determined by peak-calling software , see Methods ) and putative subtelomeric sequence were further examined for Sir4 ChIP enrichment ( an example is shown in Figure 5B ) . Sb-Sir4 associated with one end of each of these contigs and usually with an internal region as well , typically within 10 kb of the contig end . Interestingly , in the Sc-Sir4-only hybrid , Sc-Sir4 association was observed at the contigs' ends , but not at the internal regions that bound Sb-Sir4 . This result suggested that Sc-Sir4 , even in the absence of Sb-Sir4 , was capable of associating with S . bayanus telomere ends , presumably via the conserved Rap1 protein , but could not make some additional contacts necessary to associate with internal sequences . To test whether the Sc-Sir4 molecules bound to S . bayanus telomeres were capable of silencing S . bayanus subtelomeric genes , we measured the transcription of candidate subtelomeric ORFs in S . bayanus wild-type , Sb-sir4Δ::Sc-SIR4 , and Sb-sir4Δ strains . Importantly , the expression of all three putative subtelomeric genes increased in Sb-sir4Δ cells ( Figure 5C ) . Although Sc-Sir4 was capable of silencing Sb-YIR039c and an ORF located on Contig_626 , it could not repress the transcription of an ORF on Contig_511 located almost 9 kb from the main peak of Sc-Sir4 ChIP . Thus , Sc-Sir4 could bind to and silence at least a subset of S . bayanus telomeric regions . It was possible that S . bayanus had subtelomeric regulatory elements that promoted silencing , in addition to the Rap1-binding terminal repeats . Depending on the sequence of a particular element , or its proximity to the telomere end , Sc-Sir4 may or may not have been capable of binding and silencing . The cross-species complementation and ChIP analyses suggested that the incompatibility between Sc-SIR4 and Sb-HML and HMR was caused by the failure of one or more physical interactions occurring at S . bayanus silencers . In principle , the lack of productive Sc-Sir4 association with Sb-HML and Sb-HMR could have resulted either from an S . bayanus-specific inhibitor of silencing that Sc-Sir4 could not overcome or an S . bayanus-specific positive regulator of silencing ( e . g . , Sb-Rap1 or Sb-Sir1 ) with which Sc-Sir4 could not interact . To distinguish between these models , in an S . cerevisiae strain , we replaced the Sc-HMR locus with Sb-HMR containing the URA3 reporter , including the flanking silencer elements ( Figure 6A ) . If S . bayanus encoded an inhibitor of silencing that Sc-Sir4 could not overcome , Sb-HMR should be silenced in S . cerevisiae , given the strong conservation of ORC , Rap1 , and Abf1 proteins and the Rap1 and Abf1-binding sites in the HMR-E silencer [5] . If , however , Sc-Sir4 failed to be recruited to S . bayanus silencers , we would expect little or no silencing of Sb-HMR in S . cerevisiae . Upon transfer into S . cerevisiae , Sb-HMR was silenced extremely poorly ( Figure 6B , row 1 ) . However , the transplanted Sb-HMR locus could still be silenced in the context of the S . cerevisiae chromosome in hybrids made by mating the S . cerevisiae Sb-HMR strain to wild-type S . bayanus . The transplanted Sb-HMR locus was silenced to approximately the same degree as the native Sb-HMR locus in hybrids ( Figure 6B , row 2 , compare with 6C rows 1 and 2 ) . The slightly incomplete silencing of the transplanted Sb-HMR was largely due to the Sb-SIR4 dosage sensitivity observed in the original set of hybrids ( Figure 2A , row 3 ) , as silencing was strengthened in Sb-SIR4/Sb-SIR4 hybrids ( Figure 6B , row 3 ) . Thus , the lack of silencing of Sb-HMR in hybrids expressing only Sc-Sir4 ( Figure 2A , rows 2 and 5 ) was not due to an inhibitor of silencing encoded elsewhere in the S . bayanus genome . Rather , the incompatibility was encoded in the Sb-HMR locus itself , requiring S . bayanus-specific silencing proteins to interpret Sb-HMR-specific sequence information . These “interpreter” proteins potentially included DNA-binding proteins such as ORC , Rap1 , and Abf1 or proteins indirectly associated with silencers , such as Sir1 , Sir4 , or both . Alignments of Sc-HMR and Sb-HMR suggested that their functional divergence was due to changes in the silencer sequences between the two species ( Figure 7 ) . The HMRa1 gene was 83% identical between S . cerevisiae and S . bayanus ( the promoter was 93% identical ) , well above the genome-wide average of 62% identity for all intergenic regions , and the mating-type cassette-homology sequences ( shared with MAT and HML ) approached 100% identity ( Figures 6A and 7 ) . Notably , the silencer sequences share well below the genome-wide average identity for intergenic regions and are difficult to align outside of the conserved Rap1 and Abf1 sites [5] . The simplest model consistent with the results so far was that the silencing incompatibility was limited to Sir4 , with Sc-Sir4 having a more restricted range of interactions than Sb-Sir4 . To test this possibility , we replaced Sc-SIR4 with Sb-SIR4 in the S . cerevisiae strain bearing Sb-HMR . If the incompatibility involved only SIR4 and silencers , Sb-SIR4 should restore silencing to Sb-HMR . Indeed , the S . cerevisiae strain with Sb-SIR4 and Sb-HMR indeed showed a modest increase in silencing relative to the Sc-SIR4 Sb-HMR strain , confirming that changes in Sir4 itself contributed to the silencing incompatibility . However , this silencing increase—a 5-fold change—was detectable only as an increase in FOA resistance , and was still at least 100-fold below the level of HMR silencing seen in the hybrids ( Figure 8A , row 2; compare with Figure 6B , row 2 ) . Thus , although a portion of the incompatibility could be explained by SIR4 and silencer co-evolution , one or more additional S . bayanus proteins were likely required to recruit Sb-Sir4 efficiently or to stabilize its association with S . bayanus silencers . Interestingly , Sc-Sir4's very weak ability to silence the transplanted Sb-HMR locus resulted in the low-frequency appearance of FOA-resistant colonies ( occurring at an approximate frequency of 5×10−5; Figure 8A , row 1 ) . Within these colonies , which grew at nearly the rate expected of Ura- strains , the cells were able to grow under conditions that killed the majority of cells that did not form colonies . Hence this silencing occurred at low frequency , but was nonetheless heritable . Indeed , Sb-HMR silencing by either Sb-Sir4 or Sc-Sir4 was fully dependent on S . cerevisiae Sir1 ( Figure 8A ) , whose role is to promote the establishment of heritable silencing . That Sb-HMR could be silenced at all in S . cerevisiae suggested that a critical subset of Sc-ORC , Rap1 , and Abf1 bound productively to Sb-HMR silencers . It was therefore possible that providing additional S . bayanus silencing proteins could stabilize interactions between the S . cerevisiae DNA-binding proteins and S . bayanus silencers . Likely candidates to provide this presumptive function were the S . bayanus Sir1 paralogs , Kos1 , Kos2 , and Kos3 , with Kos3 being the most structurally distinct from Sir1 , yet the most similar to the ancestral member of the Sir1 family [24] . Interestingly , Sb-KOS3 enhanced Sb-HMR silencing synergistically with Sb-SIR4 , but not with Sc-SIR4 ( Figure 8B; compare rows 1 , 5 , and 10 ) . None of the other Sir1 paralogs of S . bayanus provided a dramatic enhancement of Sb-HMR silencing . The Sb-HMR Sb-SIR4 +Sb-KOS3 strain showed 100-fold better silencing than the Sb-HMR Sc-SIR4 strain ( Figure 8B , compare rows 1 and 10 ) . This result was particularly interesting because Sir4 interacts weakly and non-specifically with DNA [44] , and Kos3 is not thought to bind DNA at all . Thus , the “interpretation” of differences between the Sb-HMR and Sc-HMR silencers by Sb-Kos3 and Sb-Sir4 presumably required some sort of HMR-allele-specific collaboration with silencer-binding proteins that could be interpreted by Sb-Kos3 and Sb-Sir4 in a species-specific way . By sequence conservation , Rap1 and Abf1 binding sites can be detected in the Sb-HMR-E silencer , but the ORC binding site is not readily identified ( Figure 7 ) [5] . Given Sb-Sir4's dependence on Sir1 and Kos3 , and their dependence on ORC [20] , [45] , [46] , our results suggested two likely explanations for why Sb-HMR was not silenced in S . cerevisiae: either Sc-ORC bound S . bayanus silencers less well than S . cerevisiae silencers , or Sc-ORC bound equivalently but failed to promote silencing because it was in a suboptimal conformation or context with respect to other silencer binding proteins . In either case , the subsequent interactions with Sc-Sir1 and Sc-Sir4 might suffer . To test whether Sc-ORC indeed bound to S . bayanus silencers , we performed ChIP analysis on HA-tagged Sc-Orc5 in S . cerevisiae bearing Sb-HMR . Sc-Orc5 associated with the Sb-HMR-E silencer , albeit at a level several-fold below its association with Sc-HMR-E ( Figure 9A , left panel; note log scale on y-axis ) . A parallel analysis with Sc-Abf1 ChIP showed robust association of this protein with both Sc-HMR-E and Sb-HMR-E silencers ( Figure 9A , right panel ) . We note that both Sc-Orc5 and Sc-Abf1 associations with Sb-HMR-E showed small alterations in the Sb-SIR4 strain relative to the Sc-SIR4 strain . However , these changes did not correlate with Sb-HMR silencing levels ( Figure 8A , rows 1 and 2 ) . These ChIP data were consistent with Sb-HMR silencers having conserved functional binding sites for ORC and Abf1 . To test whether Sc-ORC , Rap1 , and Abf1 indeed participated in S . bayanus silencing , we monitored silencing of Sb-HMR in hybrids lacking either species' complement of each of these proteins ( out of the six ORC subunits , we focused on Orc1 because it directly interacts with Sir1 ) . Because RAP1 , ABF1 , and ORC1 are essential , we assayed silencing in hybrids heterozygous for each gene . S . cerevisiae diploids sensitized to detect silencing defects at HMR show strong silencing defects if either SIR1 or SIR4 dosage is also reduced [47] . Similarly , Sb-HMR silencing was weakly compromised in hybrids whereas Sc-HMR was not ( Figure 2A ) , potentially providing a sensitized background to uncover similar types of genetic interactions . For this reason , any such silencing defects in heterozygous hybrids were expected to affect silencing of Sb-HMR but not Sc-HMR . Indeed , Sb-HMR , but not Sc-HMR , was further derepressed in hybrids lacking either Sc-RAP1 or Sb-RAP1 ( Figure 9B , top panel; Figure S3 ) . Note that Sb-HMR was fully silenced in S . bayanus RAP1/rap1Δ diploids; therefore , reduced RAP1 dosage per se did not cause the loss of silencing observed in the hybrid ( Figure 9B , bottom panel ) . Thus , Sc-Rap1 participated in Sb-HMR silencing in hybrids , likely by direct binding to S . bayanus silencers . In contrast to the analysis with RAP1 , Sb-HMR was derepressed to a greater extent in hybrids lacking Sb-ORC1 but not in hybrids lacking Sc-ORC1 ( Figure 9B , top panel ) . Again , Sb-HMR was fully silenced in S . bayanus ORC1/orc1Δ diploids ( Figure 9B , bottom panel ) , ruling out simple dosage explanations . Hence , Sb-Orc1 was more important for Sb-HMR silencing in hybrids than Sc-Orc1 , suggesting that a partial species restriction existed with respect to ORC binding or activity at Sb-HMR silencers . Heterozygosity of ABF1 had no effect on either Sb-HMR or Sc-HMR silencing ( Figure 9B , top panel; Figure S3 ) . The rapid sequence and functional divergence of SIR4 between closely related species suggested that an interesting evolutionary force may have contributed to the functional divergence of this gene . To test whether a specific function of the Sir4 protein had been under positive selection within the sensu stricto clade , we aligned SIR4 coding sequence from all five species and computed the ratio of nonsynonymous to synonymous divergence ( henceforth ω , also known as dN/dS ) across the whole gene . The value of ω for SIR4 was 0 . 44 , substantially higher than the genomic average of 0 . 10 . Only 16 of 4 , 894 loci we analyzed had a higher ω , indicating that SIR4 was indeed one of the most rapidly evolving genes in the budding yeast genome . A value of ω significantly greater than 1 is evidence of positive selection [48] . Therefore , a value of 0 . 44 might suggest that the SIR4 coding region did not evolve under positive selection . However , because Sir4 is a large protein we investigated whether sub-regions or individual codons might have ω>1 . To determine whether rapidly evolving Sir4 residues might lie within known functional regions of the protein , we computed ω in 102 bp ( 34-codon ) windows throughout the SIR4 open reading frame ( Figure 10A ) . Consistent with our previous whole-gene estimate , the median ω value for all windows in SIR4 was 0 . 43 ( Figure 10A , solid horizontal line ) with a range from 0 . 02 to 1 . 87 . Because ω estimates calculated in short windows are subject to stochastic noise , we compared the results of this analysis to ∼1 , 500 , 102 bp windows drawn from other S . cerevisiae coding regions . The median of these ω values was 0 . 05 , and 95% of windows lie between 0 . 0001 and 0 . 42 ( Figure 10A , dashed lines ) . These comparisons supported two conclusions . First , because the median ω for SIR4 was comparable to the most extreme values in other genes , the unusual molecular evolution of this gene extended over a large fraction of its length . Second , the non-random distribution of windows with high ω suggested that the rapid evolution of certain residues was connected to functional changes within specific regions of the Sir4 protein . In support of this suggestion , simulations of SIR4 evolution indicated that 1 . 3% of 102 bp windows are expected to have ω>1 by chance compared to 6 . 7% observed in sensu stricto SIR4 ( unpublished data ) . The high-ω windows in SIR4 were therefore unlikely to reflect noise and instead indicated that the most rapidly evolving codons are concentrated in particular regions of SIR4 . Indeed , although the Rap1- and Sir3-binding coiled-coil domain was largely protected from the rapid evolution of SIR4 , residues within the PAD ( Partioning and Anchoring of plasmids ) and the putative N-terminal regulatory domains [37] showed striking signatures of rapid evolution ( Figure 10A ) . To provide an independent , statistically robust analysis of SIR4 evolution in this clade , we used a likelihood-ratio test to compare nested models of sequence evolution that either allowed or did not allow a subset of codons to have a value of ω>1 . The model allowing ω>1 ( M8 ) fit the data significantly better than the alternative model ( M7; p = 5×10−4 ) , indicating that some codons were likely to be evolving under positive selection ( Figure 10B ) . The posterior probability of ω≥1 exceeded 0 . 75 for 11 codons ( Figure 10A and C; ω≥1 . 5 in all cases ) , however for no single codon did the posterior probability exceed the nominal significance level of 0 . 95 . Inclusion of SIR4 sequences from species outside the sensu stricto was not possible because of poor alignment quality . In summary , although we were not able to identify specific codons that were unambiguously under positive selection , these data suggested that multiple codons within SIR4 , including some within the PAD and N-terminal regulatory domains , exhibit signatures of extremely rapid sequence evolution in the Saccharomyces sensu stricto clade . To examine whether the rapid sequence evolution of SIR4 showed a phylogenetic correlation with the functional divergence we observed , we fit models that allowed different branches of the SIR4 tree to have different values of ω . If such a correlation were observed ( i . e . , more rapid evolution of SIR4 along the S . cerevisiae lineage than along the S . bayanus lineage ) , then positive selection on a specific silencing function of SIR4 during the evolution of S . cerevisiae would be likely . Although increased estimates of ω were obtained for some branches ( notably the shared S . cerevisiae/S . paradoxus branch; ω = 0 . 55 ) , none were statistically supported , suggesting that there have been no dramatic shifts in the selection pressures operating on SIR4 since the divergence of the sensu stricto ( Figure S4 ) . We note that a change in selection pressure that affected only a subset of codons could easily have gone undetected .
The Sir4 protein and silencers diverged rapidly in concert , a process that was accompanied by loss of three Sir1 paralogs in the S . cerevisiae lineage [24] . As silencing was robustly maintained in each species , it was likely that these factors had co-evolved such that coding changes in Sir4 and a reduction in Sir1 family members led to compensatory changes in silencers , or vice versa . The asymmetrical complementation of SIR4 alleles ( Figure 2A ) , and the enhanced ability of Sb-Sir4 to bind S . cerevisiae silent loci compared to its own silent loci ( Figure 4B ) , suggested that S . cerevisiae silencer elements had become stronger than those of S . bayanus , while S . cerevisiae Sir1 and Sir4 proteins had become weaker ( operationally defined ) than S . bayanus Sir4 and its four Sir1 paralogs . The intra-species combinations of Sir1 and Sir4 proteins and silencers allowed efficient nucleation of silencing complexes at HML and HMR in each species . Broadly speaking , we imagine two possible evolutionary paths for this co-evolution , with variations on either path possible . In an “adaptive” model , hypothetical selective pressure ( s ) induced coding changes in Sir4 and reduction in Sir1 family members ( Zill et al . in preparation ) , which then required “strengthening” mutations ( for example , a change that increased the affinity of ORC for a silencer ) in the silencers to maintain robust silencing . In a “constructive neutral” model [49] , strengthening mutations accumulated in silencers at random , thus relaxing the selective constraints to maintain Sir1 paralogs and certain Sir4 residues . Once Sir1 paralogs were lost , the “stronger” silencers would need to be maintained by purifying selection . Our evolutionary analyses using the PAML software supported a role for positive selection acting on multiple sites within SIR4 during the evolution of the sensu stricto species ( Figure 10 ) . However , the SIR4 gene showed an unusually high evolutionary rate across most of its length . Furthermore , although we identified 11 rapidly evolving residues that may suggest specific regions' contributions to the functional divergence of Sir4 , none of these crossed the 0 . 95 threshold for statistical significance . ( We note that 4 of the 11 fastest evolving sites were localized within the PAD domain , which mediates interactions between Sir4 and Esc1 at the nuclear periphery , and between Sir4 and the Ty5 retrotransposon integrase protein . Interestingly , Esc1 is also one of the most rapidly diverging proteins in Saccharomyces species [O . Zill , unpublished observations] . ) Extensive directed mutational analyses will be necessary to test whether the sites under selection are responsible for the functional divergence between Sc-Sir4 and Sb-Sir4 . An important question relevant to these models is , in which lineage did the observed changes in Sir4 and silencer function occur relative to the common ancestor of S . cerevisiae and S . bayanus ? Although accurate determination of the ancestral state of the silencing mechanism will require extensive evolutionary analyses , it appears that S . bayanus has retained at least two ancestral characters that S . cerevisiae has lost . First , Kos3 , the ancestral Sir1-related protein , has been lost in S . cerevisiae . Second , the SIR4 gene from K . lactis , an outgroup to the Saccharomyces clade , was able to complement silencing function in S . bayanus sir4Δ mutants ( Zill et al . in preparation ) . That a Sir4 protein from a species outside of Saccharomyces is compatible with S . bayanus silencers suggests that these elements did not “gain” a restrictive property in the S . bayanus lineage . The more likely scenario is that Sir4 changed in the S . cerevisiae lineage such that its range of interactions with other species' silencers has become restricted , consistent with earlier observations of cross-species function of Sir4 [50] . It will therefore be of interest to understand in detail the mechanism of silencing in S . bayanus and to determine what forces caused the dramatic shift in Sir1 and Sir4 functionality in the S . cerevisiae lineage . We measured the rates of SIR4 evolution ( ω ) along all branches in the sensu stricto clade but did not observe a notable asymmetry in these rates ( Figure S4 ) . Thus , the functional asymmetry between Sc-Sir4 and Sb-Sir4 was probably localized to a few sites and may not be related to the broad evolutionary forces that have acted on SIR4 across all five species in this clade . Perhaps the most striking finding of this study was that the heterochromatin proteins that showed the most dramatic evidence of co-evolution with silencers , Sir1 and Sir4 , were not the ones that bind specific DNA sites . Rather , these proteins associate with DNA indirectly via the conserved regulatory proteins Rap1 , Abf1 , and ORC ( Figure 11 ) . The key evidence demonstrating functional co-evolution between Sir4 and the Sir1 family and silencers came from attempts to reconstitute Sb-HMR silencing in S . cerevisiae . The changes in Sir4 sequence were not sufficient to explain the inability of Sc-Sir4 to function at S . bayanus silencers: expression of Sb-Sir4 in an S . cerevisiae strain was only modestly effective in silencing an Sb-HMR locus transplanted into that strain ( Figure 8A ) . The Sir1-dependence of the rare , but heritable , silencing events mediated by Sb-Sir4 at Sb-HMR in S . cerevisiae suggested that the limitation involved proteins dedicated to establishing silencing . Indeed , adding Sb-Kos3 , the ancestral member of the Sir1 family , together with Sb-Sir4 enhanced silencing of Sb-HMR in S . cerevisiae by 100-fold ( Figure 8B ) , although not completely . It was possible that the site-specific DNA-binding proteins ORC , Rap1 , and Abf1 had also co-evolved with silencer sequences . If this were the case , we would expect hybrids lacking the Sb-ORC , Sb-Rap1 , or Sb-Abf1 proteins to have shown defective Sb-HMR silencing . However , only Sb-Orc1 inactivation ( and by inference , inactivation of the entire Sb-ORC complex ) showed the expected S . bayanus allele-specific effect on Sb-HMR silencing ( Figure 9B ) . This effect of Sb-ORC1 deletion on Sb-HMR silencing was relatively modest , and the addition of Sb-ORC1 ( together with Sb-SIR4 ) had no effect on Sb-HMR silencing in S . cerevisiae reconstitution experiments ( unpublished data ) . Because Sc-Orc1 , Sc-Rap1 , and Sc-Abf1 were capable of supporting Sb-HMR silencing in hybrids ( Figure 9B ) and in S . cerevisiae ( Figure 8B ) , their DNA-binding domains' interactions with silencers were largely conserved across species and hence were not engaged in notable co-evolution with silencers or with Sir4 . Indeed , we were able to ChIP Sc-Orc5 and Sc-Abf1 on the Sb-HMR-E silencer in S . cerevisiae ( Figure 9A ) . Together , these results suggested that the cis-acting differences between the two species' silencers were interpreted largely indirectly , via interactions between ORC , Sir1/Kos3 , and Sir4 , with a somewhat lesser contribution of differences in ORC-silencer DNA interactions . Why did Sc-Sir4 not bind efficiently to S . bayanus silencers ? Simple explanations such as sequence divergence between S . cerevisiae and S . bayanus silencers precluding sequence-specific contacts with Sir4 are unlikely because biochemical data on Sir4 point to a lack of sequence-specific binding to DNA [44] . Instead , Sir4 is recruited to silencers predominantly via protein-protein interactions [20] , [37] , [51] . It is unlikely that different proteins bind the silencers in the two species as the preponderance of evidence points to ORC , Rap1 , and Abf1 as the critical silencer-binding proteins in both species ( Figures 5 , 7 , and 9 ) . Further , the residues mediating Sc-Orc1 interaction with Sc-Sir1 [52] , [53] are conserved in Sb-Orc1 ( our unpublished observations ) . Hence we are forced to consider models in which something special about how ORC , Rap1 , and Abf1 bind S . bayanus silencers prevents Sc-Sir4 from interacting with Rap1 or creates a requirement for specific interactions with the Sir1 paralogs that can only be made by Sb-Sir4 . Perhaps the precise juxtaposition or conformation of these site-specific DNA-binding proteins allow or restrict interactions with a particular species of Sir4 . Alternatively , perhaps a reduced affinity of S . bayanus silencers for ORC or Rap1 , or the ensemble of nucleation proteins , is compensated by binding energy provided by Sb-Kos3 ( and possibly additional Sir1 paralogs ) and Sb-Sir4 , but not by Sc-Sir4 . Indeed , complete silencing of both Sb-HML and Sb-HMR requires Sb-Sir1 , Sb-Kos1 , and Sb-Kos2 [24] . Therefore , it is possible that a relatively weak binding site ( such as for Rap1 ) in the Sb-HMR-E silencer could be compensated by increased binding energy provided to the nucleation complex in trans by the combination of Sb-Sir4 and the Sir1 paralogs . Additionally , we note that a requirement for multivalent interactions may help explain why Sir4 fails to interact stably with Rap1 at the many Rap1 binding sites throughout the genome . An unexpected finding of the Sir4 comparative ChIP-Seq experiment provided insight into the mechanism of silent chromatin assembly . The Sb-Sir4-assisted Sc-Sir4 incorporation into Sb-HML and HMR ( Figure 4A ) suggested two distinct types of interactions made by Sir4 proteins at these loci: only Sb-Sir4 was capable of making stable contacts either with the Sir1 paralogs , or perhaps with Rap1 . However , as Sc-Sir4 was capable of mediating telomeric silencing at some S . bayanus telomeres ( Figure 5C ) , it appeared that Sc-Sir4 could interact productively with the Sb-Rap1 protein . In addition , there was a second and qualitatively distinct mode of Sir4 protein association that was species-independent but occurred only if the species-specific interaction occurred . Three types of interactions might account for the secondary mode of Sc-Sir4 association with Sb-HML and Sb-HMR: direct Sb-Sir4-Sc-Sir4 interaction via a conserved dimerization surface [38] , Sc-Sir4 binding to Rap1 via the conserved C-terminal coiled-coil domain , or Sc-Sir4 interaction with deacetylated histone tails [18] . We note that Sc-Sir4 association with the Sb-HMR-E silencer increased in the presence of Sb-Sir4 at least as much as did its association with internal regions of Sb-HMR ( Figure 4A ) . Thus , this secondary mode of Sir4 interaction did not appear to be restricted to regions of Sb-HMR where the deacetylated histones reside ( silencers are nucleosome-free regions ) . Further studies will resolve whether Sb-Sir4-assisted Sc-Sir4 incorporation involves contacts with multiple silencing proteins versus simple Sir4-Sir4 dimerization and whether it requires Sir2 catalytic activity . Additionally , the enhanced interaction of Sb-Sir4 across Y′ elements at S . cerevisiae telomeres ( Figure 4B ) suggested that novel or changed interactions in the hybrids somehow led to enhanced Sir4 occupancy of these regions . This differential long-range occupancy by Sir complexes presents an opportunity to ask whether Sir1 and Sir4-mediated interactions during Sir complex nucleation regulate the “strength” of silent chromatin over a distance . Alternatively , Sb-Sir4 ( and potentially other S . bayanus silencing proteins ) may have been less sensitive to factors that exclude Sc-Sir4 from the Y′ elements . The species-specific Sir4 distributions occurring in these interspecies hybrids should be further dissected to understand the determinants limiting silent chromatin formation across subtelomeric regions . Another unusual property of the interspecies hybrids led to a weak silencing defect affecting Sb-HMR but not Sc-HMR ( Figure 2A , row 1; Figure 3B ) . In hybrids lacking Sc-Sir4 this defect was more evident ( Figure 2A , row 3 ) , which paradoxically suggested that Sc-Sir4 protected Sb-HMR silencing in the presence of Sb-Sir4 , despite having no ability to silence Sb-HMR on its own . How might Sc-Sir4 have “enhanced” Sb-Sir4 function at S . bayanus silent loci in hybrids ? Strong evidence compatible with Sc-Sir4 protecting Sb-Sir4 from being titrated by Sc-specific sequences was the ability of both Sb-Sir4 and Sc-Sir4 to bind extensively to S . cerevisiae telomeres , as described in the Results . Hence , the hybrid state may result in a dosage sensitivity to Sb-SIR4 not evident in S . bayanus SIR4/sir4Δ intra-species diploids due to additional binding sites provided by the Sc-X and Y′ elements , and potentially other elements . We note the resemblance of this “Sb-Sir4 sequestration” model to the “Circe effect” proposed to explain Sc-Sir4-mediated clustering of S . cerevisiae telomeres [54] . Gregor Mendel's studies were motivated by a desire to understand the emergent properties of interspecies hybrids , such as hybrid vigor , that were of great practical significance at the time . Although he became famously distracted by discovering two fundamental laws of genetics , his original interest in the processes by which hybrid species are not necessarily the “average” of the two parental species remains as interesting today as it was practically important in Mendel's day . Indeed , the striking asymmetry in the ability of Sb-Sir4 to silence Sc-HMR , but inability of Sc-Sir4 to silence Sb-HMR ( Figure 2A ) , was the seminal observation that inspired this study . By and large , however , in interspecies hybrids of S . cerevisiae and S . bayanus , a protein from either species was fully capable of providing all of that protein's function to hybrids . Although this result could be anticipated from the ability to “clone by complementation” genes of one species by their function in another , this study established that symmetry of complementation is an important general consideration . For example , the essential proteins Rap1 and Abf1 from either species had all the functions necessary to support viability of the hybrids , and we established that Sir2 and Sir3 of both species were fully interchangeable ( Figure S1 ) , despite being members of a complex in which another member of that complex , Sir4 , has extraordinary divergence . By extrapolation , asymmetrical deviations from a general expectation of cross-species compatibility , such as in the case of Sir4 , may signal situations of uncommon interest . The studies presented here capitalized on the extraordinary genetic properties of interspecies hybrids to tease out important dimensions to the evolution and structure of silent chromatin in yeast . Although silencing behavior in these yeast hybrids was rather unusual , some type of defect might have been anticipated from recent studies of hybrid sterility or lethality genes in Drosophila , which have implicated rapidly evolving heterochromatin proteins as key factors contributing to interspecies genetic incompatibility [55] , [56] . There is presently no reason to believe that SIR1 or SIR4 play roles in the post-zygotic genetic incompatibility between budding yeast species . It is notable , however , that in budding yeast multiple regulatory sites mediating silencing have rapidly evolved in a phylogenetically asymmetrical fashion along with a set of divergent silencing proteins , paralleling observations of rapid evolution in Drosophila heterochromatin [55] , [57] , [58] . It will be of great interest to determine whether the similar patterns of heterochromatin evolution in these distant taxa reflect similar underlying evolutionary processes . The unprecedented resolution of Sir4 distribution provided by ChIP-Seq methods calls into question earlier models for silenced chromatin assembly , and in particular the so-called mechanism of spreading ( reviewed in [7] , [59] ) . In the common view , Sir protein recruitment to the silencers or to telomeres allows the deacetylation of H4K16-Ac on adjacent histones , creating new binding sites for additional Sir protein complexes , with sequential cycles of deacetylation and binding leading to spreading of Sir-protein complexes across all nucleosomes in silenced chromatin . The strikingly uneven distribution of Sir4 at HML and HMR ( Figure 4A , note y-axes ) , and at the telomeres ( Figure 4B ) , as shown here and in K . lactis [60] , is not entirely inconsistent with the common view of heterochromatin spreading , but is in no way anticipated by it . Clearly , high-resolution characterization of all Sir proteins by ChIP-Seq has the potential to force substantial revision or replacement of the current view .
All S . cerevisiae strains were of the W303 background . Generation of marked S . bayanus strains from type strain CBS 7001 has been described [61] . All yeast strains were cultured at 25°C in standard yeast media . One-step gene replacement and C-terminal 13xMyc tag integration have been described previously [36] , [62] , and these genetic manipulations were performed identically for S . bayanus , S . cerevisiae , and S . cerevisiae/S . bayanus hybrids . The HMR::URA3 reporter strains were constructed independently in S . cerevisiae sir4Δ and S . bayanus sir4Δ haploid strains , wherein the HMRa1 ORF was replaced with the K . lactis URA3 ORF by PCR-based gene targeting , leaving the HMRa1 promoter intact . For most experiments , interspecies hybrids were made by crossing S . bayanus MATα HMR::URA3 strains ( wild-type , sir4Δ , or SIR4-13xMyc ) to S . cerevisiae MATa strains ( wild-type , sir4Δ , or SIR4-13xMyc ) . For ORC1 , RAP1 , and ABF1 heterozygote analysis , gene targeting was performed directly in hybrid diploids or S . bayanus diploids . Three independent transformants were analyzed in all cases . Sc-SIR4-13xMyc and Sb-SIR4-13xMyc alleles were shown to be functional by two independent silencing assays in each case: by mating ability in S . cerevisiae SIR4-13xMyc and S . bayanus SIR4-13xMyc haploid strains and by FOA resistance in hybrid diploids bearing the appropriate HMR::URA3 reporter ( unpublished data ) . The Sc:: ( Sb-HMR::URA3 ) replacement allele ( Figure 6A ) was generated in two steps . The Sb-HMR::URA3 cassette plus 1 kb of leftward-flanking sequence was PCR-amplified out of the S . bayanus genome , and the PCR product was used to replace the syntenic portion of Sc-HMR ( including the E silencer ) in S . cerevisiae sir4Δ strains . A HygMX marker was then targeted into the S . bayanus genome 3 kb to the right of Sb-HMR . The entire rightward-flanking 3 kb region plus the HygMX marker was PCR-amplified out of the S . bayanus genome , and the PCR product was used to replace the syntenic portion of Sc-HMR in the S . cerevisiae genome ( including the I silencer ) . The Sc:: ( Sb-HMR::URA3 ) replacement allele therefore included a total of 5 . 5 kb of Sb-HMR sequence , plus the 1 . 7 kb HygMX marker . To construct the SIR4 replacement alleles , the Sc-SIR4 and Sb-SIR4 genes were separately cloned into the yeast plasmid pRS315 [63] such that the LEU2 marker was 5′ of , and in opposite orientation to , each SIR4 gene . Each SIR4 gene plus the LEU2 marker was PCR-amplified from each plasmid . The LEU2-Sc-SIR4 PCR product was used to replace the URA3 marker at the Sb-SIR4 locus in an S . bayanus sir4Δ::URA3 leu2 strain; likewise , the LEU2-Sb-SIR4 PCR product was targeted into the Sc-SIR4 locus in an S . cerevisiae sir4Δ::URA3 leu2 strain . The integrated Sc-SIR4 gene was shown to silence Sc-HMR::URA3 in hybrids ( Figure 3A ) , and the integrated Sb-SIR4 gene was shown to silence Sb-HMR::URA3 in hybrids ( Figure 6C ) and Sc-HML and Sc-HMR in S . cerevisiae strains ( Zill et al . in preparation ) . The expression level of each SIR4 replacement allele was determined by quantitative RT-PCR ( Figure S2A ) . Assays of yeast strain growth on FOA and CSM/-Ura media were performed using standard “frogging” techniques . Briefly , for each strain , a 10-fold dilution series of yeast cells at an approximate density of 4×107/mL was spotted onto each plate . For Figures 2 , 3A , 6 , and 8A plates were photographed after 2 d for YPD , and after 3 d for FOA and CSM/-Ura . For Figure 8B , plates were photographed after 3 d for all media . For Figure 9B , plates were photographed after 3 d for FOA and YPD , and after 5 d for CSM/-Ura . We note that some changes in silencing could be seen only on FOA and not on CSM/-Ura . Incomplete silencing of the HMRa1 promoter likely led to heterogeneous expression states within the population of cells , with some remaining silent while others were expressed [64] . RNA isolation was performed using the hot-phenol method [65] . Total RNA was digested with Amplification grade DNase I ( Invitrogen ) and purified using the RNeasy MinElute kit ( Qiagen ) . cDNA was synthesized using the SuperScript III First-Strand Synthesis System for RT-PCR and oligo ( dT ) primer ( Invitrogen ) . Quantitative PCR on cDNA was performed using an MX3000P machine ( Stratagene ) and the DyNAmo HS SYBR Green qPCR kit ( NEB ) . Amplification values for all primer sets were normalized to actin ( ACT1 ) or SEN1 cDNA amplification values . Samples were analyzed in triplicate from three independent RNA preparations . Yeast whole cell extracts were prepared using 20% TCA and solubilized in SDS loading buffer plus 100 mM Tris base . SDS-PAGE and immunoblotting were performed using standard procedures and the LiCOR imaging system . Anti-c-Myc antibody from rabbit ( Sigma , Cat . No . C3956 ) was used to detect Myc-tagged Sir4 and Abf1 proteins . Mouse anti-Pgk1 antibody ( Invitrogen , Cat . No . 459250 ) was used to verify equal loading . The S . cerevisiae Orc5-HA strain derivation has been described [66] . All chromatin immunoprecipitations ( Sir4-Myc , Orc5-HA , Abf1-Myc ) were performed as described [67] , using formaldehyde cross-linking of log phase cultures for 1 h at room temperature . IPs were performed overnight at 4°C using Anti-c-Myc-Agarose ( Sigma , Cat . No . A7470 ) and Anti-HA-Agarose ( Sigma , Cat . No . A2095 ) . Quantitative PCR was performed as described above . Orthologous S . cerevisiae and S . bayanus genes were identified on the basis of sequence similarity and syntenic context ( D . Scannell and M . B . Eisen , unpublished ) . Percent identities between 4 , 981 orthologous S . cerevisiae and S . bayanus proteins were then obtained by running BLASTP [73] with default parameters , imposing an E-value cutoff of 1×10−5 , harvesting percent identities for each HSP and calculating a length-weighted average . This will necessarily lead to some underestimation of the true divergence between protein pairs , but it is unlikely that the rank order of divergences among pairs would be significantly affected . For PAML and sliding window analyses , protein alignments were produced using FSA with default parameters [74] , and DNA alignments were obtained by back translation with RevTrans [75] . All site and branch models were fit using codeml in the PAML package [76] . To test for positive selection we compared model M8 to M7 using a χ2 test with two degrees-of-freedom . Posterior probabilities of ω>1 for individual codons were obtained from the Bayes Empirical Bayes output of M8 only . For the sliding window ( dN/dS , or ω ) analyses , a window size of 102 bp and a step-size of 3 bp were used . Only alignment windows without gaps were analyzed . For each window we used codeml to estimate a single ω using model M0 implemented in codeml . All other parameters were estimated from the data . We estimated the level of selective constraint operating on SIR4 on each branch of the Saccharomyces sensu stricto phylogeny by computing branch-specific ratios of non-synonymous to synonymous substitutions ( dN/dS , or ω ) . Briefly , we performed protein-space alignments of orthologous SIR4 coding sequences with FSA [74] and then used codeml in the PAML package [76] to fit a “free-ratio” model ( model = 1 , NSSites = 0 ) to the alignment and obtain independent estimates of ω for each branch . | As eukaryotic species evolve , transcriptionally silent portions of their genomes—termed “heterochromatin”—mutate rapidly . To maintain the “off” state of certain genes in silenced regions , regulatory DNA sequences called silencers , which reside within a rapidly mutating region , must co-evolve with the regulatory proteins that bind these sequences to turn off transcription . Although hypothesized to occur widely in nature , such “molecular co-evolution” of genetic regulators has been demonstrated in only a few cases . Unlike previous examples of gene regulatory co-evolution , we found that the transcription factors that bind silencers in two budding yeast species are , in fact , functionally interchangeable , even though the silencers are not . Surprisingly , the Sir1 and Sir4 silencing proteins , which are heterochromatin components that bind the transcription factors rather than the silencer DNA sequences per se , are the proteins engaged in rapid co-evolution with the silencers . Silencer sequences therefore contain additional , evolutionarily labile information directing the assembly of heterochromatin . As mutations in Sir1 and Sir4 over evolutionary time can compensate for changes in the silencers , this “extra information” likely involves cooperative assembly of the transcription factors with the Sir1 and Sir4 “adaptor” proteins . The localization patterns of two species' Sir4 proteins across both species' genomes in interspecies yeast hybrids illuminate unexpected features of heterochromatin structure and assembly . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/gene",
"expression",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"molecular",
"biology/molecular",
"evolution",
"evolutionary",
"biology/genomics",
"genetics",
"and",
"genomics/gene",
"function",
"genetics",
"and",
"genomics/epigenetics",
"molecular",
"biology/chromatin",
"structure"
] | 2010 | Co-Evolution of Transcriptional Silencing Proteins and the DNA Elements Specifying Their Assembly |
There is a critical need to better use existing antibiotics due to the urgent threat of antibiotic resistant bacteria coupled with the reduced effort in developing new antibiotics . β-lactam antibiotics represent one of the most commonly used classes of antibiotics to treat a broad spectrum of Gram-positive and -negative bacterial pathogens . However , the rise of extended spectrum β-lactamase ( ESBL ) producing bacteria has limited the use of β-lactams . Due to the concern of complex drug responses , many β-lactams are typically ruled out if ESBL-producing pathogens are detected , even if these pathogens test as susceptible to some β-lactams . Using quantitative modeling , we show that β-lactams could still effectively treat pathogens producing low or moderate levels of ESBLs when administered properly . We further develop a metric to guide the design of a dosing protocol to optimize treatment efficiency for any antibiotic-pathogen combination . Ultimately , optimized dosing protocols could allow reintroduction of a repertoire of first-line antibiotics with improved treatment outcomes and preserve last-resort antibiotics .
Bacteria eventually develop resistance to all antibiotics they encounter [1–3] . Unfortunately , the evolution of antibiotic resistant bacteria is accelerating due to the widespread use of antibiotics [4 , 5] . As the antibiotic pipeline is drying up and the threat of antibiotic resistance is becoming more urgent [6 , 7] , it is critical that we better utilize the antibiotics already on the market [8–10] . One of the largest and most commonly used classes of antibiotics for treating both Gram-positive and Gram-negative bacteria is the β-lactams [11–13] . Many β-lactams , such as penicillin V , amoxicillin , and first-generation cephalosporins , are first-line antibiotics; they are recommended for initial therapy because they are highly effective against non-resistant pathogens , have a lower risk of side effects , and are less expensive , relative to second-line antibiotics [14–16] . However , the rapid emergence of extended spectrum β-lactamase ( ESBL ) producing pathogens has greatly limited the use of β-lactam antibiotics [13 , 17] . ESBL-producing pathogens have significant adverse effects on clinical outcomes due to their ability to hydrolyze penicillins , broad-spectrum cephalosporins , and monobactams [6 , 18 , 19] . Patients infected with ESBL-producing pathogens have worse prognoses and , if given the incorrect treatment , mortality rates of 42–100% greater than patients receiving the correct treatment [18 , 20] . Additionally , β-lactams could promote horizontal gene transfer of virulence factors [21] and could be responsible for the spread of ESBL genes . As a precaution , most first-line β-lactams are ruled out if ESBL-producing pathogens are detected , even for ESBL-producing pathogens that appear to be sensitive to a particular β-lactam [22–25] . This is done largely out of concern for complicating drug responses that have been observed in vitro , such as the inoculum effect , a phenomenon in which the minimum inhibitory concentration ( MIC ) of an antibiotic increases as the bacterial density increases [24 , 26–30] . With first-line β-lactams ruled out , second-line antibiotics , such as carbapenems , fluoroquinolones , β-lactam/β-lactamase inhibitor combinations , glycopeptides , and cephamycins , are typically administered [31] . Although this practice is based on a valid concern , it has limitations . Specifically , second-line antibiotics are associated with higher costs and more adverse effects [32–37] . Additionally , the more frequently bacteria are exposed to second-line antibiotics , the faster the pathogens are likely to develop resistance to our last resort antibiotics [2 , 5] . Given the dearth of new antibiotics entering the market and the limited number of effective antibiotics already available , we cannot afford to disregard potentially effective antibiotics . First-line β-lactams could represent a missed opportunity for treating pathogens producing moderate levels of ESBLs . Individual bacteria that produce moderate levels of ESBL can remain sensitive to the antibiotic due to insufficient production or activity of ESBL; however , if enough bacteria are present , then the population’s collective ESBL concentration will be sufficient to render the population resistant to the antibiotic [38 , 39] . In other words , a low density population of moderate ESBL producers would lyse entirely because its collective ESBL concentration would be insufficient to inactivate the β-lactam , while a high density population would only experience partial lysis before its collective ESBL concentration can inactivate the β-lactam and promote the recovery of the surviving bacteria . This collective population recovery is time dependent [40] . Shortly after the antibiotic is first applied , the population will be reduced due to lysis and appear susceptible because it will not have yet benefited from the activity of ESBLs . Ideally , a treatment could pinpoint the time window when the most lysis has occurred and the least benefit has been experienced . Extensive studies have been carried out to devise methods to optimize treatment efficacy of antibiotics by changing the dosing period and amplitude . These studies typically examine which metric ( s ) can capture the pharmacokinetic/pharmacodynamics ( PK/PD ) of an antibiotic and be used to predict antibiotic efficacy [41–44] . Current metrics adopted in the clinical setting , such as the MIC , do not account for the time course of antimicrobial activity and are not sufficiently predictive of treatment efficacy [22 , 45–47] . Therefore , there is a need for a simple metric that characterizes this pathogen-antibiotic interaction that can be easily measured and used to design dosing protocols that will effectively clear an infection . Here , we use quantitative modeling to demonstrate a strategy for customizing regimens for a particular bacteria and antibiotic combination without needing to know the full mechanistic basis for the bacteria-antibiotic interaction . Specifically , we focus on optimizing a dosing protocol to enable β-lactams to effectively treat a moderate ESBL-producing pathogen . To help guide the design of effective protocols , we develop a metric , the recovery time , which is easy to measure and quantifies the pathogen-antibiotic interaction . Even though we assumed specific molecular mechanisms underlying this collective antibiotic response , our model illustrates that the predictive power of the recovery time is maintained for different specific molecular mechanisms and for different initial conditions . Through optimized antibiotic regimens , our strategy could extend the use of first-line antibiotics , improve treatment outcome , and preserve last-resort antibiotics .
We developed a kinetic model comprising a set of ordinary differential equations ( ODEs ) to capture the population dynamics of collectively tolerant , ESBL-producing bacteria being treated by a β-lactam ( S1 Text ) [40] . We further nondimensionalized the model to facilitate analysis . In this model , introduction of the antibiotic inhibits bacterial growth and causes lysis . β-lactamase ( Bla ) is naturally found in the periplasm of Gram-negative bacteria , where it can benefit the host bacterium by hydrolyzing the β-lactams that diffuse into the periplasm [48] . However , moderate amounts of periplasmic Bla are insufficient to protect a bacterium from high concentrations of antibiotic [38 , 49] . Conversely , sufficient amounts of Bla can accumulate to protect a population if enough bacteria are initially present . With a dense enough population , the collective intracellular and extracellular Bla , due to lysis or leaky secretion [50] , will be sufficient to degrade the antibiotic to a sublethal concentration before all cells are eliminated ( Fig . 1A ) . Thus , the survival of the population depends on establishing a collective antibiotic tolerance ( CAT ) [30] . In general , Bla expression can be constitutive or inducible by the antibiotic [51–54] . Here , we focus on constitutive Bla expression , which is most clinically relevant to the pathogens that express plasmid-mediated ESBLs [39 , 55] . However , our conclusions also apply to the case where Bla expression is inducible . They will likely apply to bacterial responses to other antibiotics if the antibiotic causes an initial decline in the population density by killing a subpopulation of cells and the population can recover when the antibiotic is subsequently degraded by an enzyme produced by the cells ( whether or not the enzyme is released into the culture ) . Using physiologically relevant parameters , our model generates PK/PD profiles that are characteristic of Bla-mediated CAT . Starting from a sufficiently high initial density , the population exhibits an initial decline upon antibiotic treatment , followed by eventual recovery due to intrinsic and Bla-mediated degradation of the antibiotic ( Fig . 1B ) . Sufficient time is needed to observe this apparent drug tolerance . If examined shortly after antibiotic treatment , the population will have just experienced significant lysis and will appear susceptible because the effects of Bla have not yet been fully recognized . For a fixed initial antibiotic concentration , the model predicts a switch-like dependence of population survival over the initial population density: the population can only survive if starting at a sufficiently high density ( Fig . 1C ) . If too few bacteria are present , the total expression of Bla from the entire population is insufficient to degrade the antibiotic fast enough to allow the population to recover . If enough bacteria are present , however , the population can endure the initial crash in density for a longer period . As such , some bacteria remain when the antibiotic concentration decreases sufficiently , due to Bla-mediated degradation , to allow the population to recover . The density-dependent survival of the population is the defining feature of the inoculum effect [28 , 56] . Our results illustrate the defining features of a CAT bacterial response involving antibiotic-triggered death . In particular , the population will appear resistant when its initial density is sufficiently high and it is given enough time to recover . These features form the basis for the preemptive practice of disregarding β-lactams when an ESBL-pathogen is identified . However , our model also indicates that the population is sensitive when its initial density is sufficiently low or when it is examined in a short time window . Given these properties , we reason that optimal antibiotic dosing may remain effective in eliminating bacteria . If so , an immediate next question is how to best design the treatment protocol . This task would be straightforward if we could determine the specific molecular mechanisms and defining parameters for each pathogen-antibiotic pair: under such a scenario , we could in theory use a model specific to the pair to examine efficacy of different dosing protocols . This is impractical , however , as many ESBL pathogens are poorly characterized at the molecular level and there are many different ESBL enzymes [57] . A more practical option would be to identify an easy-to-measure , lumped metric based on a bacterial population’s response to a single dose of antibiotic that will allow us to reliably predict its response to periodic antibiotic treatment without needing to know the underlying molecular-level parameters . A typical metric to quantify efficacy of an antibiotic is the minimum inhibitory concentration ( MIC ) , which can be measured by disk diffusion and microbroth dilution methods after a certain duration of antibiotic treatment [58] . However , the MIC measured at a particular time point does not capture the rich temporal dynamics of bacterial responses due to antibiotic-triggered death . Instead , we propose to use another lumped metric: the recovery time; specifically , this defines the time it takes a population to return to its initial density after being exposed to a dose of antibiotic ( Fig . 2 ) . By definition , the recovery time captures the dominant dynamic features of bacterial temporal response . As such , it may be a more predictive metric for the long-term outcome of periodic antibiotic treatment . We first tested the predictive power of the recovery time in injection-based dosing protocols . With the base-parameter set , our model predicts a monotonic dependence of the recovery time on the antibiotic concentrations for single-dose treatment ( Fig . 3A ) . Once the initial antibiotic concentration is high enough to cause cell lysis ( a0 > 0 . 5 ) , the recovery time increases exponentially with the initial antibiotic concentration until the antibiotic concentration is too high ( a0 > 10 ) and the recovery time becomes infinite . This dependence is an intrinsic property of antibiotic-mediated lysis . Under low concentrations of antibiotic ( 0 . 5 < a0 < 10 ) , the recovery time is primarily determined by how fast the antibiotic is degraded by Bla . Under increasing concentrations of antibiotic ( a0 > 10 ) , the rate of antibiotic degradation is essentially saturated ( limited by the population size and the constant production rate of Bla ) and the recovery time is primarily determined by the lysis rate . β-lactams’ killing rate is time- , not dose- , dependent and is reflected in the model’s lysis rate’s non-linear dependence on the antibiotic concentration ( Hill coefficient = 3 ) [59] . Once the antibiotic concentration is high enough , further increasing the concentration does not increase the lysis rate . As noted above , the recovery time could represent a simple , yet reliable , metric in predicting outcomes from periodic treatment . To test this notion , we examined the consequence of periodic dosing of varying antibiotic concentrations . For each concentration , we varied the dosing periods from 0 . 1 to 2 times the corresponding recovery time , and obtained the final population density after 100 doses . Our modeling results confirmed the predictive power of the recovery time: as long as the initial antibiotic concentration is sufficiently high to cause significant initial lysis , the population will reach a high final density if the period is greater than the recovery time; the population goes extinct otherwise ( Fig . 3B ) . Of the regimens leading to eventual population extinction ( period < recovery time ) , different combinations of antibiotic concentrations and dosing periods eliminate a population with varying efficacy . To quantify this efficacy , we calculated the minimum number of doses necessary to reduce the population density to below 10-10 ( Fig . 3C ) . The resulting landscape shows a strong dependence on antibiotic concentration and the corresponding recovery time . When the antibiotic concentration is too low and the recovery time is close to 0 , the number of doses required to clear the infection is very large , regardless of the dosing frequency . When the antibiotic concentration is very high and the corresponding recovery times approach infinite , then the number of doses is very low . However , there is an intermediate range of antibiotics with intermediate recovery times that show variation in the number of doses necessary to clear the infection . Concentrations producing the longer recovery times require fewer doses because they can reduce the bacterial density more severely than concentrations with shorter recovery times . For intermediate antibiotic concentrations ( 1 < a0 < 10 ) to be most effective , the model suggests they should be delivered at low-to-intermediate period lengths ( period = 20–50% recovery time ) at which the population is most vulnerable . At the end of each period , the bacteria are still lysing , have almost reached minimum density , but have not yet experienced the benefits of Bla . At this point , the antibiotic has not been completely removed; thus the population will be subjected to a slightly higher concentration of antibiotic at each additional dose . If the antibiotic is delivered too frequently , the accumulated antibiotic increases the rate of lysis , thus causing higher amounts of Bla to be released , ultimately leading to the faster removal of the antibiotic . However , Bla cannot fully degrade the antibiotic before the next dose is added and the population quickly dies off . Although the population is cleared , a higher number of doses is necessary because the degree of lysis per dose is not maximized . In other words , subsequent doses are applied before the full extent of lysis from the previous dose is observed . However , if the antibiotic is delivered too infrequently , then the population will have the chance to recover between doses . Once again , these conditions do not maximize the degree of lysis per dose and more doses are necessary to achieve the same amount of population decrease associated with doses applied more frequently . A final aspect to consider when designing a regimen is the total amount of antibiotic delivered ( Fig . 3D ) . Although some of the model’s regimens using higher concentrations of antibiotic ( a0 > 10 ) are associated with fewer doses , they have the highest net antibiotic concentration . These concentrations may not be optimal , due to potential adverse effects associated with using excessive amounts of antibiotic , such as the destruction of the normal microbial flora , interference with the immune response , increased nephrotoxicity , and selection for antibiotic resistant mutants [32 , 60–63] . Also , efficient use of antibiotics can help reduce treatment cost [14 , 35] . Using dose number and total antibiotic delivered , an effective and realistic regimen can be designed by minimizing the number of doses , the delivery frequency , and the total antibiotic delivered . We note that the predictive power of the recovery time is maintained for low or moderate inoculum sizes . In particular , our modeling demonstrates that a multi-dose regimen will clear a population if the time between doses is less than one recovery time , regardless of effective antibiotic concentration and inoculum size ( S1 Fig ) . Similar to the base case , the regimen can be optimized to have the fewest doses and the lowest net antibiotic concentration delivered by selecting the lowest concentration of antibiotic associated with the longest recovery time . Additionally , the predictive power of the recovery time is maintained for an antibiotic with dose-dependent killing ( Hill coefficient = 1 ) or an antibiotic with time-dependent killing ( Hill coefficient = 10 ) : a multi-dose regimen will clear a population if the time between doses is less than one recovery time , regardless of effective antibiotic concentration and degree of antibiotic-mediated killing ( S2 Fig ) . The predictive power of the recovery time can be applied to bacteria with varying rates of Bla synthesis and accumulation as long as the antibiotic concentration applied has an effective recovery time ( S3 Fig A-T ) . When the bacteria are producing and accumulating Bla at a very fast rate ( S3 Fig P-T ) , most individual bacteria can sufficiently protect themselves ( CAT is no longer necessary ) and the population experiences little or no decline in density . Consequentially , the model predicts that effective treatment protocols would shift to higher antibiotic concentrations capable of inducing significant lysis in more resistant bacteria . The predictive power is upheld as long as the recovery times associated with subsequent doses are sufficiently similar to the original recovery time measured from a single dose . Recovery times of subsequent doses depend on two main factors: the activity of Bla in the environment and the concentration of antibiotic . On one hand , if there is insufficient time for Bla to degrade between doses , then it will compound with each dose until the population is being protected by higher concentrations of Bla , relative to when the first dose was administered . As a result , the increasing pool of Bla will degrade the antibiotic faster , the recovery time of subsequent doses will decrease , and the population can recover when dosed at period lengths less than the original recovery time ( S4 Fig A-B ) . This would happen in scenarios where the antibiotic concentration applied is insufficient to counterbalance the Bla that is either expressed at high levels or has an increased rate for hydrolyzing an antibiotic . The loss of predictive power in this case can be avoided by using a sufficiently strong antibiotic concentration . On the other hand , if there is insufficient Bla to degrade the antibiotic between doses , then the antibiotic will compound with each dose until the population is being exposed to higher concentrations of antibiotic , relative to when the first dose was administered . As a result , the increasing concentration of antibiotic will kill more cells , the recovery time of subsequent doses will increase , and the population will not be able to recover when dosed with period lengths equal to the original recovery time ( S4 Fig C-D ) . Many antibiotics , such as β-lactams , are most effective when applied continuously for long periods of time [64 , 65] . Thus , we also modeled the predictive power of the recovery time in intravenous ( IV ) -drip based protocols , where a set concentration of antibiotic is delivered over a set duration during each dosing period . Here , we delivered the antibiotic dose over three time units and measured the corresponding recovery time ( Fig . 4A ) . Similar to the injection recovery times , the IV-drip recovery times increase monotonically as the concentration of the dose increases , more Bla is required to remove the antibiotic , and more of the population lyses . In contrast , the IV-drip therapy has a narrower range of intermediate antibiotics with 0 < recovery time < 100 . Some of the lower concentrations that are effective for injection treatment ( 0 . 5 < a0 < 1 ) are ineffective for IV drip treatment because the dose is too weak when delivered over a longer period of time . However , when the dose concentration is sufficiently high , the IV-drip recovery time is longer than the injection recovery time because the IV-drip is exposing the bacteria to a higher concentration for a longer period of time ( Fig . 4B ) . Again , we use the recovery time from a single IV dose to establish the range of dosing frequencies able to eliminate the bacterial population . At each dosing concentration ( for a fixed time duration of 3 ) , we applied 100 doses of the antibiotic at periods ranging from the infusion duration ( τ = 3 ) to 2 times the corresponding recovery time and calculated the resulting final bacterial density . The model shows that the predictive power of the recovery time is maintained when the antibiotic dosing concentration is sufficiently large with a long enough recovery time ( a0 > 1 . 5 ) : a multi-IV-dose regimen will eventually eliminate the population if the dosing period is less than one recovery time , regardless of effective antibiotic concentration ( Fig . 4C ) . There is slight deviation from this for a0 < 1 . 5 due to the corresponding recovery times being too short for the Bla to be reduced to a baseline concentration before the next round of lysis and Bla release occurs . As a result , periods less than one recovery time could fail to eradicate the infection because the Bla concentration compounds with each subsequent dose , the antibiotic is degraded more quickly , fewer cells lyse , and the population can recover ( S4 Fig ) . Similar to the injection based therapy , the IV-drip reduced a population constitutively producing high concentrations of Bla as long as the period was less than one recovery time and the initial antibiotic concentration was sufficiently high to cause significant initial decline ( S3 Fig U-Y ) . However , the IV-drip protocols retained a larger range of effective antibiotic concentrations than the injection based protocols . This robustness is due to the antibiotic concentration continuously being replenished from the IV-drip . If a high enough concentration is maintained for sufficient time , the population’s Bla concentration will not be able to remove the antibiotic fast enough to prevent lysis and the population will decrease with each subsequent round of IV-drip infusion . Thus , these results suggest that IV-drip based regimens could serve as a platform for effectively applying first-class β-lactams to clear constitutive producers of high levels of ESBLs . We next evaluated the efficacy of each effective concentration-period combination by calculating the minimum number of doses necessary to reduce the population density to below 10-10 ( Fig . 4D ) . Relative to the injection protocol , the IV-drip therapy has a narrower region of intermediate dose numbers , reflecting the narrow region of intermediate recovery times . Similarly to the injection based regimens , the intermediate antibiotic concentrations ( 1 < a0 < 5 ) require the least doses when delivered at low-to-intermediate period lengths ( period = 20–60% recovery time ) because that is when the population is most vulnerable . Again , the initial antibiotic concentrations too low to have a recovery time ( a0 < 1 ) do not clear the infection , regardless of the dosing interval or number of doses applied . The concentrations with an infinite recovery time ( a0 > 5 ) require only a single dose and thus the dosing frequency does not matter . Although the number of doses necessary to clear an infection might be the same for a range of antibiotic concentrations and periods , the least amount of total antibiotic is needed for intermediate antibiotic concentrations applied at 20–60% of the associated recovery time ( Fig . 4E ) . A bacterial population often consists of phenotypically or genetically heterogeneous subpopulations[66 , 67] . For instance , different cells may express different levels of Bla , have different growth rates , or exhibit different sensitivities to the same antibiotic . This heterogeneity could compromise the predictive power of the recovery time . To examine this notion , we extended our injection-based model to account for two cases , each dealing with a mixture of two subpopulations ( S1 Text ) . In one case , one subpopulation grows much more slowly and exhibits much greater tolerance to the antibiotic . In the other , two subpopulations display different degrees of collective antibiotic tolerance .
Most antibiotic regimens are based on empirical observations of how bacterial infections responded to an antibiotic [32 , 78 , 79] . However , these regimens may be suboptimal both because they were not initially designed to handle resistant bacteria and because the current diagnostic assays cannot accurately predict how resistant pathogens will respond to them . It is critical that we develop a new strategy for using the existing antibiotics more effectively or our medical care will return to a state equivalent to that of a pre-antibiotic era . Ideally , the new strategies would be based on the molecular mechanisms underlying antibiotic resistance . However , this is impractical , given that many pathogens’ resistance mechanisms have not been characterized and they evolve rapidly . To this end , we propose using the recovery time as a lumped metric that can characterize a pathogen’s response to an antibiotic without requiring knowledge of the underlying mechanism . We used a kinetic model to test the ability of the recovery time to predict ESBL-producing pathogens’ responses to periodic dosing of β-lactams . Our simulation results suggest that the recovery time of a single dose can be used to design optimal multi-dose regimens for multiple delivery methods , including injections and continuous IV drip , various inoculum sizes , bacteria with a range of Bla production levels , and certain heterogeneous populations . Optimal dosing regimens for treating Bla-producing bacteria with a β-lactam would apply intermediate concentrations of antibiotic that have long recovery times at time intervals corresponding with when the bacterial density has been minimized . Furthermore , our modeling results suggest that regimens using lower , yet still lethal , concentrations of antibiotic can be as effective as regimens using higher concentrations . Reducing the amount of antibiotic the host is exposed to may be important to minimize the perturbation of the host’s microbiota and other defense mechanisms [32 , 60–62 , 80] , which could have long-lasting detrimental effects . Also , under certain conditions , a higher concentration of antibiotic can lead to selection of more resistant subpopulation of bacterial pathogens [81] . Although this model considers the population level response to an antibiotic , there is a significant amount of gene-expression noise at the single cell level [66 , 67] . If an antibiotic were applied such that the population would have the chance to recover between doses , then the antibiotic would select for the bacteria expressing higher levels of resistance genes ( as demonstrated in S6 Fig F ) . Ultimately , this would direct the evolution of the population towards an inherently more resistant infection than before the antibiotic treatment was applied . Our proposed method would minimize this problem by delivering subsequent doses of antibiotic before a more resistant population grew to a significant density . The recovery time of a pathogen under a single dose of antibiotic is a metric that is easy to measure and could guide the choice of an appropriate multi-dose antibiotic regimen for a wide range of infections . Measurements of the recovery time can be carried out in high resolution using commercially available microplate readers [82] . A critical step entails the construction of a comprehensive recovery time database for various pathogens under different antibiotics ( Fig . 5 ) . When a new bacterial pathogen is identified , its recovery times to a range of antibiotic concentrations will be recorded in vitro for different starting densities . Based on these measurements , regimens with varied concentrations and period lengths will be tested for different inoculum sizes . From these results , the period length , dose number , and antibiotic concentration can be optimized for a particular pathogen in vitro . Before entering this information into the database , the PK/PD of the particular antibiotic will be necessary to determine the concentration of antibiotic that should be delivered such that the concentration at the site of infection matches the concentration selected from the in vitro experiments . Given this database , a proper diagnosis of a pathogen , and an estimate of the severity of the infection ( e . g . inoculum size ) , one can readily identify the scenarios in which first- and second-line antibiotics may still be applied and chose the most effective treatment protocol . Whenever a new pathogen arises , it can be evaluated and added to the library . The ability to predict the outcome of a multi-dose treatment without knowing the underlying resistant mechanism would remove the uncertainty that prevents clinicians from using first-line β-lactams when an ESBL-producing pathogen is detected . Given ESBL-producing bacteria’s prevalence [19 , 83–85] , our proposed strategy could help to minimize the rate at which these bacteria develop resistance to more extreme antibiotics by ensuring that we do not overlook effective first-line antibiotics before moving on to more extreme antibiotics .
The interaction between a β-lactam and a bacterial population expressing Bla can be simplified to the interactions between three main components: population density ( n ) , antibiotic concentration ( a ) , and Bla concentration ( b ) . Our base model consists of the following ordinary differential equations: dndτ= ( g−l ) n ( 1 ) dboutdτ=lbin*−γ2bout−κIV ( τ ) bout ( 2 ) da dτ=κIV ( τ ) ainject− ( bout+αbin* ) ( a1+a ) −γ3a−κIV ( τ ) a ( 3 ) g= ( 1−n ) ( σ1σ1+a ) ( 4 ) l=γ1 ( aHσ2H+aH ) ( σ4σ4+bin ) ( 5 ) bin=κ ( rg+γ4 ) ( 6 ) bin*=βnbin ( 7 ) r=aσ3+a ( 8 ) where g and l represent bacteria growth and lysis , respectively . Initial conditions of n ( 0 ) = 0 . 1 , b ( 0 ) = 0 , and a ( 0 ) = 0 . 01–100 were used for all simulation results , except for S2 Fig where n ( 0 ) = 0 . 01 or 0 . 001 . The rest of the parameters are defined in a table in S1 Text . See S1 Text for further details of the model development and for the extended models that account for heterogeneous populations . Minor modifications are introduced to account for the IV drip protocol or dynamics of a mixture consisting of two subpopulations ( S1 Text ) . | Antibiotic resistance is a growing problem that the World Health Organization describes as “one of the top three threats to global health . ” To date , bacteria have developed resistance to all antibiotics used in clinical settings . Unfortunately , the evolution of antibiotic resistant bacteria is accelerating , as antibiotics continue to be misused and overused . As the antibiotic pipeline is drying up , it becomes increasingly critical to utilize the antibiotics already on the market more effectively . The key to designing better regimens lies in the ability to predict how bacteria will respond to a particular antibiotic treatment . Because of this , we need a simple metric that characterizes this pathogen-antibiotic interaction that can be easily measured and used to design dosing protocols that will effectively clear an infection . To help guide the design of effective protocols , we use quantitative modeling to develop a metric that is easy to measure and quantifies the pathogen-antibiotic interaction . Through optimized antibiotic regimens , our strategy could extend the use of first-line antibiotics , improve treatment outcome , and preserve last-resort antibiotics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Bacterial Temporal Dynamics Enable Optimal Design of Antibiotic Treatment |
Tether proteins attach the endoplasmic reticulum ( ER ) to other cellular membranes , thereby creating contact sites that are proposed to form platforms for regulating lipid homeostasis and facilitating non-vesicular lipid exchange . Sterols are synthesized in the ER and transported by non-vesicular mechanisms to the plasma membrane ( PM ) , where they represent almost half of all PM lipids and contribute critically to the barrier function of the PM . To determine whether contact sites are important for both sterol exchange between the ER and PM and intermembrane regulation of lipid metabolism , we generated Δ-super-tether ( Δ-s-tether ) yeast cells that lack six previously identified tethering proteins ( yeast extended synatotagmin [E-Syt] , vesicle-associated membrane protein [VAMP]-associated protein [VAP] , and TMEM16-anoctamin homologues ) as well as the presumptive tether Ice2 . Despite the lack of ER-PM contacts in these cells , ER-PM sterol exchange is robust , indicating that the sterol transport machinery is either absent from or not uniquely located at contact sites . Unexpectedly , we found that the transport of exogenously supplied sterol to the ER occurs more slowly in Δ-s-tether cells than in wild-type ( WT ) cells . We pinpointed this defect to changes in sterol organization and transbilayer movement within the PM bilayer caused by phospholipid dysregulation , evinced by changes in the abundance and organization of PM lipids . Indeed , deletion of either OSH4 , which encodes a sterol/phosphatidylinositol-4-phosphate ( PI4P ) exchange protein , or SAC1 , which encodes a PI4P phosphatase , caused synthetic lethality in Δ-s-tether cells due to disruptions in redundant PI4P and phospholipid regulatory pathways . The growth defect of Δ-s-tether cells was rescued with an artificial "ER-PM staple , " a tether assembled from unrelated non-yeast protein domains , indicating that endogenous tether proteins have nonspecific bridging functions . Finally , we discovered that sterols play a role in regulating ER-PM contact site formation . In sterol-depleted cells , levels of the yeast E-Syt tether Tcb3 were induced and ER-PM contact increased dramatically . These results support a model in which ER-PM contact sites provide a nexus for coordinating the complex interrelationship between sterols , sphingolipids , and phospholipids that maintain PM composition and integrity .
Most lipids are synthesized in the endoplasmic reticulum ( ER ) and distributed to other membranes by non-vesicular mechanisms . These mechanisms act in conjunction with lipid metabolic networks to maintain the unique lipid profile of the plasma membrane ( PM ) and subcellular organelles , and enable rapid membrane lipid remodeling in response to signals and stresses [1–3] . An attractive hypothesis is that non-vesicular lipid transport and lipid biosynthetic and regulatory pathways intersect at ER-PM membrane contact sites ( MCSs ) , where protein tethers retain the ER and PM within about 15–60 nm of each other [4–9] . In this view , ER-PM MCSs would serve as a nexus , coordinating requirements in the PM for lipids with their production in the ER [3 , 9] . How this coordination is accomplished is not well understood . Here , we report on the interplay between sterol and phospholipid homeostasis at ER-PM MCSs . Cholesterol—and its yeast counterpart ergosterol—are synthesized in the ER and transported by non-vesicular mechanisms to the PM [10 , 11] , where they are found at high concentrations corresponding to about 40 mole percent of PM lipids , i . e . , one out of every two to three lipids in the PM is a sterol . The spontaneous exchange of sterols between membranes is slow in vitro and undetectable in vivo , primarily because sterol desorption from the membrane is energetically expensive [12–14] . To move sterols efficiently between the ER , PM , and other membranes , cells make use of sterol transport proteins ( STPs ) , whose proposed function is mainly to reduce the energy barrier for sterol desorption , thereby extracting sterols into a binding pocket within the protein for transit through the cytoplasm [12] . STPs may operate freely in the cytoplasm or at MCSs . Soluble and membrane-bound STPs might work in parallel to provide redundant mechanisms for sterol exchange . As transport is predicted to be rate-limited by the desorption step rather than diffusion of the STP–sterol complex through the cytoplasm [12] , the proximity of the ER to the PM at an MCS may not determine the sterol transport rate unless STPs are restricted to these sites . Because a number of sterol biosynthetic enzymes are enriched in PM-associated ER membrane fractions [4] , it is attractive to consider that the biosynthetic and transport machineries may colocalize to ER-PM MCSs , effectively channeling sterol between compartments [9] to facilitate sterol homeostasis . The identity of STPs is controversial and the role of MCSs in sterol transport is unexplored . STP candidates in yeast include members of two protein families: soluble Osh proteins ( related to mammalian oxysterol-binding protein [OSBP] [15] ) and membrane-bound lipid transfer proteins anchored at MCSs ( Lam ) ( members of the StARkin superfamily of steroidogenic acute regulatory [StAR] protein–related lipid transfer [StART] proteins [16 , 17] ) . Osh4 binds sterols and phosphatidylinositol-4-phosphate ( PI4P ) [18] and by toggling between its sterol and PI4P bound states , it has been shown to transport sterol against a concentration gradient between vesicle populations in vitro [19] . While this activity may account for certain aspects of sterol homeostasis , Osh4 is not required for retrograde sterol transport [20] , nor is it essential for the high rate of sterol exchange between the ER and PM , as evinced by robust sterol transport in cells where all seven OSH genes are inactivated ( oshΔ ) [21] . The seven Osh proteins share overlapping essential activities [22] , but because Osh6 and several other Osh proteins do not bind sterols [23–25] , sterol transport is not a function shared by the entire family . Lam proteins each have one or two sterol-binding StARkin domains . The purified domains have been shown to catalyze sterol exchange between vesicles in vitro [26–28] . Lam1–Lam4 are integral ER membrane proteins located at the cell cortex , where they might function as sterol transporters , similar to the mammalian StARkin STARD3 , which is anchored to endosomal membranes and has been suggested to facilitate endosome–ER cholesterol transfer [29] . Although elimination of Lam proteins does not inhibit the bidirectional transport of newly synthesized ergosterol between the ER and PM , sterol organization at the PM is altered [30] . If Osh and Lam proteins catalyze ER-PM sterol transport , then they must do so redundantly with each other and/or with additional STPs yet to be identified . Our results address this point . In addition to their proposed role in sterol homeostasis , ER-PM MCSs are known to be involved in phospholipid biosynthesis and turnover . Phosphatidylcholine ( PC ) is synthesized via Cho2 and Opi3-mediated methylation of phosphatidylethanolamine ( PE ) . Opi3 is an ER-localized membrane protein in yeast that has been proposed to act in trans at ER-PM MCSs to convert PM-localized PE to PC [31] . In the absence of Opi3 function ( either through lack of the enzyme or disruption of ER-PM MCSs ) , cells rely on the Kennedy pathway , through which PC is synthesized from choline taken up from the growth medium . It has been proposed that phosphoinositide turnover also occurs at ER-PM MCSs , where the ER-localized PI4P phosphatase Sac1 may act in trans to turn over PI4P synthesized in the PM [5] . These examples highlight the possibility that ER-PM MCSs may contribute to a wide range of reactions that underlie cellular phospholipid homeostasis . In yeast , about 45% of the PM retains a closely associated cortical ER ( cER ) membrane [5 , 32] , and this association requires a number of tethering proteins that staple the ER and PM together [5 , 31 , 33–35] . Six ER-PM tethering proteins are currently known ( Fig 1A ) : the three tricalbins ( Tcb1–3 ) , which are yeast homologues of the extended synaptotagmin ( E-Syt ) family of membrane tethers; Ist2 , a member of the TMEM16-anoctamin family of ion channels and phospholipid scramblases; and the yeast vesicle-associated membrane protein ( VAMP ) -associated protein ( VAP ) homologues Scs2 and Scs22 . Several of these tethers appear to be Ca2+ regulated in mammalian cells , and from their embedded location within the ER membrane , a number of them make contact with the PM through associations with phosphoinositides and/or other phospholipids [6 , 8 , 36–40] . By eliminating all six of these tethering proteins , Manford and colleagues [5] created Δtether yeast cells , in which large sections of the PM are devoid of cortically associated ER membrane . However , as these authors noted , additional unknown tethers must still exist in Δtether cells , given that small regions of cER were still associated with the PM [5] . Consistent with this conclusion , the localization of Lam1–Lam4 close to the PM near presumptive MCSs is unaffected in Δtether cells [30] . These results suggest that elimination of the six tether proteins is not sufficient to remove all ER-PM contacts and that additional proteins/mechanisms must exist to account for the remaining ER-PM association . In order to eliminate residual cER in Δtether cells , we focused on Ice2 , an integral ER membrane protein ( Fig 1A ) with established roles in cER inheritance [33 , 44] , ER-PM contact [31 , 33] , phospholipid synthesis from stored neutral lipid [31 , 45] , and ER quality control [46] . Ice2 was first shown to facilitate cER redistribution and inheritance along the PM from mother cells into daughter buds [44] . In cells lacking both ICE2 and SCS2 , cER association at the PM is disrupted more than for each single mutant [31 , 33] . The defect in ER-PM membrane association in scs2Δice2Δ cells is linked to dysfunctional PC synthesis , likely because the Opi3 methyltransferase is no longer able to act on its PM-localized lipid substrate in trans at contact sites [31] . When cells enter stationary phase , Ice2 has another function , in which it generates a bridge between the ER and lipid droplets [45] . This membrane attachment has been proposed to play a role in channeling droplet-generated diacylglycerol ( DAG ) to the ER for phospholipid synthesis when cells resume growth [45] . Curiously , the ER-associated degradation ( ERAD ) substrate carboxypeptidase Y* ( CPY* ) is stabilized in ice2Δ cells compared with wild-type ( WT ) cells , pointing to a direct or indirect role for Ice2 in ER-associated degradation [46] . We reasoned that because of its various ER functions , specifically including the generation of ER-PM contacts during mitosis , Ice2 might account for the residual cER in Δtether cells . If ER-PM contact is necessary for non-vesicular sterol transfer , the rate of ER-PM sterol exchange and/or PM sterol organization would be inhibited by the elimination of all ER-PM MCSs . Likewise , if MCSs serve as regulatory interfaces to coordinate pathways for phospholipid metabolism in the ER and PM , then removing ER-PM MCSs would be predicted to alter cellular phospholipid profiles . We now report that disruption of ICE2 in Δtether cells sharply reduces ER-PM associations to the predicted frequency of randomly finding untethered ER in the vicinity of the PM . The availability of these Δ-super-tether ( Δ-s-tether ) cells now permits direct tests of hypotheses concerning how ER-PM MCSs impact non-vesicular sterol exchange and inter-membrane lipid regulation . We now report that the bidirectional movement of sterols between the ER and PM is unaffected in Δ-s-tether cells , indicating clearly that the sterol transfer machinery in yeast is either absent from or not uniquely localized to ER-PM MCSs . Nonetheless , sterol pools within the PM bilayer of Δ-s-tether cells are dramatically altered , and the rate of transbilayer sterol movement within the PM is slowed . We discovered that these defects were associated with changes in the organization and composition of PM lipids and could be largely reversed by supplementing cells with choline or by expressing a nonspecific artificial ER-PM tether . Phospholipid dysregulation in the PM was revealed by changes in the levels of sphingolipids and other PM lipids , as well as by the accumulation of PI4P at the PM of mother Δ-s-tether cells . Interestingly , Δ-s-tether cells were inviable when they also lacked Osh4 or Sac1 . After testing the associated roles of Osh6 and the ER-membrane association of Sac1 , we conclude that Osh4 and ER-PM MCSs are redundant regulators of PI4P and phospholipid homeostasis . Finally , we found that ER-PM MCS formation is responsive to cellular sterol levels , whereby the tether protein Tcb3 is induced in sterol-depleted cells , resulting in a dramatic increase in membrane association . These results suggest that ER-PM contact sites are dynamic interfaces that adjust and respond to lipid metabolism to maintain PM composition and organization .
Despite the dramatic reduction in ER-PM contact sites caused by the elimination of six tether proteins , the extent of cER in Δtether cells [5] is both significant and heterogeneous , with >35% of the cells possessing fluorescently labeled ER in the vicinity of the PM ( Fig 1B and 1C ) and individual cells displaying as much as 20% of the average cER found in WT cells ( Fig 1D and 1E and S1A Fig ) . Residual cER in Δtether cells is also evinced by the cortical localization of green fluorescent protein ( GFP ) -tagged Ysp2/Lam2/Ltc4 ( hereafter called Lam2; S2 Fig ) and other Lam proteins [30] . These observations suggest that there are additional mechanisms for generating ER-PM association [8] . Because the gene encoding the ER membrane protein Ice2 ( Fig 1A ) has a negative genetic interaction with SCS2 , and Ice2 plays roles in maintaining cER structure and mediating the inheritance of cER from mother cells into daughter buds [33 , 44] , we hypothesized that Ice2 may contribute to ER-PM association and that its presence in Δtether cells could account for the residual cER seen in these cells . To explore this possibility , we used confocal fluorescence microscopy to determine the subcellular distribution of Ice2-GFP in Δtether cells ( S3 Fig ) . Ice2-GFP was observed throughout the ER , including cytoplasmic strands radiating from nuclear ER towards the cell periphery ( S3A Fig ) . Notably , Ice2-GFP fluorescence was observed at the cell cortex , visualized by co-expressing the PM marker red fluorescent protein ( RFP ) -Ras2 ( S3A Fig ) . Optical sections focused at the cell surface showed a concentration of Ice2-GFP fluorescence at remaining ER cortical spots that were visualized using the fluorescent pan-ER marker RFP-ER ( dsRED-SCS2220–244 ) ( S3B Fig ) . This localization pattern resembles that of Scs2 and differs from that of the Tcbs and Ist2 that are restricted to ER-PM MCS spots [5 , 35 , 42 , 47] . These data suggest that Ice2 is correctly localized to contribute to ER-PM tethering . To eliminate residual cER in Δtether cells , we therefore deleted ICE2 in tandem with the other Δtether mutations . We predicted that the resulting Δ-s-tether cells would be largely devoid of ER-PM contact sites and this was indeed the case . We confirmed the near absence of PM-associated cER in Δ-s-tether cells as follows . First , expression of the pan-ER marker RFP-ER revealed that , unlike WT and Δtether cells , in which cortical fluorescence was observed in 100% and >35% of cells , respectively , fewer than 10% of Δ-s-tether cells had fluorescence at the cell cortex ( Fig 1B and 1C ) . Whereas cortical fluorescence in WT cells and many Δtether cells occurred in the form of linear strands running parallel to the cell perimeter in equatorial views ( Fig 1B , arrowheads ) , the occasional fluorescence seen at the cortex of a small fraction ( <10% ) of Δ-s-tether cells was in the form of punctae , possibly corresponding to the ends of ER tubules or coincidental positioning of the ER near the PM in the focal plane chosen for imaging ( Fig 1B , arrows ) . Second , whereas GFP-Lam2 is localized exclusively in cortical punctae in WT and Δtether cells ( about 15 cortical punctae per cell , on average , for both strains ) , cortical expression of Lam2 is considerably reduced in Δ-s-tether cells ( about four cortical punctae per cell , on average; S2A and S2B Fig ) , even though the expression level of the protein is unaffected ( S2C Fig ) . Third , cER association along the PM in Δ-s-tether cells was all but absent , as quantified by measuring the cER/PM length ratio in equatorial views of individual cells obtained by transmission electron microscopy ( Fig 1D ) . The average cER/PM ratio was 0 . 48 and 0 . 04 in WT and Δtether cells , respectively [5] , but only 0 . 017 in Δ-s-tether cells ( Fig 1E ) . The decrease in the cER/PM ratio in Δ-s-tether versus Δtether cells was statistically significant ( Fig 1E , right panel ) , representing not only an approximately 60% lower average value but also a considerable tightening of the distribution of cER/PM values ( Fig 1E and S1A Fig ) . Finally , we generated 3D models of WT and Δ-s-tether cells by reconstructing images obtained with a focused ion beam–scanning electron microscope ( FIB-SEM ) . These models ( Fig 1F ) illustrate that the extensive cER coverage of the PM in WT cells is clearly absent in Δ-s-tether cells , where a spaghetti-like accumulation of cytoplasmic tubular ER is observed instead . We estimate that the low amount of cER in Δ-s-tether cells ( Fig 1E and S1B Fig ) can be accounted for by the random chance of finding untethered ER at the cortex ( Materials and methods ) . Δ-s-tether cells grow normally on rich media but poorly on minimal media ( Fig 2B and 2E ) , suggesting that the lack of ER-PM contact sites disrupts cell metabolism . If this were indeed the case , then an artificial ER-PM tethering protein might allow the cells to grow normally . Several of the natural ER-PM tethers , e . g . , Tcb1–3 and Ist2 ( Fig 1A ) , have a modular architecture , and this design principle was used to assemble an artificial tether ( "ER-PM staple" ) from unrelated non-yeast proteins . As building blocks for the ER-PM staple , we used ( i ) two ER-anchoring Trans-membrane domains from herpes virus mK3 E3 ubiquitin ligase , ( ii ) extended helices from mammalian mitofusin 2 to span the gap between the PM and ER , and ( iii ) the C-terminal polybasic region from mammalian Rit1 ( RitC ) that targets the PM ( Fig 2A ) . We fused GFP to the N-terminus in order to visualize the ER-PM staple in cells . Expression of the ER-PM staple from the yeast actin promoter largely rescued the growth defect of Δ-s-tether cells cultured on solid medium ( Fig 2B ) , indicating that the artificial staple is a functional substitute for the endogenous tether proteins . Fluorescence microscopy revealed that the ER-PM staple localizes to cER in both WT and Δ-s-tether cells , consistent with the idea that it generates ER-PM contact sites , albeit fewer than endogenous tethers ( Fig 2C and 2D ) . The overall distribution of the ER-PM staples was similar in WT and Δ-s-tether cells , although the staples in Δ-s-tether cells aggregated in larger spots with greater fluorescence . The finding that a wholly heterologous construct can replace endogenous tether proteins in rescuing the poor growth of Δ-s-tether cells indicates importantly that the tethers ( Fig 1A ) perform a nonspecific bridging function relevant to cell growth that is exclusive of any tether-specific activities . A further conclusion from this result is that the proposed lipid transfer function of the synaptotagmin-like mitochondrial-lipid-binding protein ( SMP ) domains of the Tcbs ( Fig 1A ) [48] is not required for cell growth , consistent with observations in HeLa cells and mice lacking E-Syts [49 , 50] . Why would the absence of ER-PM tethers cause cells to grow slowly , with growth rescue being achieved by an artificial tether ? Tavassoli and colleagues [31] suggested that the ER-anchored phospholipid methyltransferase Opi3 acts at ER-PM contact sites in trans to generate a pool of PC at the PM that is necessary for growth . If Opi3 is unable to act on the PM , as would be expected for cells lacking ER-PM contact sites , then a choline supplement must be provided to generate the necessary PC via the Kennedy pathway [31] . Indeed , we found that Δ-s-tether cells achieve normal growth when the medium is supplemented with choline ( Fig 2E ) . Importantly , the extent of cER was not detectably different in choline-grown Δ-s-tether cells ( Fig 2F ) . We conclude that choline supplementation bypasses the requirement for ER-PM contact sites to support cell growth . Consistent with these findings , the deletion of either OPI3 or CHO2 in Δ-s-tether cells severely exacerbated the growth defect of the cells unless choline was provided ( S4A Fig ) . However , in the absence of choline , Opi3 overexpression effectively suppressed the choline-dependent growth defect of Δ-s-tether cells ( S4B Fig ) , potentially by providing an alternative route to generate PC pools at the PM . Unlike choline supplementation to the growth medium , the addition of ethanolamine or inositol , which promote PE and PI synthesis , respectively , did not rescue Δ-s-tether growth defects ( S5 Fig ) . The ability of choline to rescue the poor growth of Δ-s-tether cells , and the functional requirements of Δ-s-tether cells for CHO2 and OPI3 , suggested that PC synthesis/levels might be dysregulated in these cells . Surprisingly , whole cell lipidomics ( Fig 2G ) revealed that PC levels were only about 20% lower in Δ-s-tether cells compared with WT cells , but the relative amounts of a number of other lipids , notably PE , phosphatidylserine ( PS ) , and the yeast sphingolipids inositol-phosphoceramide ( IPC ) and mannosylinositol phosphoceramide ( MIPC ) , were considerably reduced . Reduced levels of these lipids were also found in Δtether cells that grow almost as well as WT cells , but the lipid compositional effects were generally more pronounced in Δ-s-tether cells . For example , we found the mole percentage of IPC content in Δtether cells to be about 67% of that in WT cells , whereas in Δ-s-tether cells , the level of this lipid fell to about 40% of that in WT cells . Increases in some lipids were also measured , most notably DAG , which was 1 . 3-fold higher in Δ-s-tether compared with WT cells . These results suggest a possible threshold effect , in which the lipid compositional changes in Δ-s-tether cells have a severe impact on growth , whereas the somewhat lesser changes in Δtether cells do not . Comparison of the lipid composition of Δ-s-tether cells cultured with or without choline supplementation revealed changes that could be predicted based on the deployment of the Kennedy pathway because of the availability of choline ( S6A Fig ) . Thus , PC and PS levels increased in the choline-supplemented cells , bringing the levels of these lipids closer to those in WT , whereas levels of mono- and dimethyl-PE ( mPE and mmPE , respectively ) fell . Choline supplementation of Δ-s-tether cells also resulted in an increase in MIPC levels , although other sphingolipids and their precursors were only slightly affected . Unlike lipid compositional changes seen upon choline addition , rescue of Δ-s-tether growth defects by the artificial tether indicated a different mechanism . The lipidomic profile of Δ-s-tether cells expressing the artificial tether was more consistent with a restoration of normal phospholipid synthesis through the cytidine diphosphate diacylglycerol ( CDP-DAG ) pathway ( S6B Fig ) . The artificial tether increased PS , PI , and mmPE levels , although levels of PC were not appreciably changed from Δ-s-tether cells cultured without choline . The artificial tether also affected storage lipids: esterified ergosterol and triacylglycerol ( TG ) showed especially large increases as a proportion compared to WT . To our surprise , expression of the artificial tether in Δ-s-tether cells did not restore levels of sphingolipids or their immediate precursors . We conclude that the molecular basis of growth rescue in Δ-s-tether cells by choline and the artificial tether is multifactorial and is likely finely tuned to the precise pools and relative abundance of several lipids , depending on the mode of suppression . To determine whether ER-PM contact sites play a role in sterol exchange between the two membranes , we compared the rate of retrograde transport of dehydroergosterol ( DHE ) from the PM to the ER in WT , Δtether , and Δ-s-tether cells using a previously described assay ( Fig 3A ) [21 , 51] . DHE is a fluorescent sterol that is widely used as a reporter of intracellular sterol transport and distribution [52]; it is particularly appropriate as a sterol reporter in yeast cells , as it is closely related to ergosterol and as effective as ergosterol in supporting the growth of hem1Δ cells that cannot synthesize sterols [21] . To load DHE into the PM , the cells are incubated under hypoxic conditions to overcome "aerobic sterol exclusion , " which represses endogenous sterol synthesis in favor of exogenous sterol import . When the DHE-loaded cells are transferred to aerobic conditions , ergosterol synthesis resumes and DHE is displaced from the PM . On reaching the ER , DHE becomes esterified by ER-localized sterol acyltransferases . The extent of esterification—detected by the appearance of lipid droplets containing fluorescent DHE or direct measurement of DHE esters by high-performance liquid chromatography ( HPLC ) analysis of lipid extracts from the cells [21]—provides a measure of retrograde transport . At the start of the chase period , DHE fluorescence was observed as a “ring stain” in WT , Δtether , and Δ-s-tether cells ( Fig 3B ) , indicating insertion of the fluorescent sterol into the PM [21 , 53] . After a 2 h incubation ( “chase” ) under aerobic conditions , fluorescence was concentrated in lipid droplets in WT and Δtether cells , but the same punctate fluorescence was not observed in Δ-s-tether cells ( Fig 3B ) . To quantify retrograde transport , the amount of imported DHE that was converted into DHE esters was measured at different times following the aerobic chase ( Fig 3C ) . DHE esterification proceeds linearly after a lag period of about 1 h , during which the cells adapt to aerobic conditions , allowing resumption of ergosterol synthesis [21] . Compared to WT or Δtether cells , we observed an approximately 4-fold decrease in the rate of transport-coupled esterification of DHE in Δ-s-tether cells ( Fig 3C and 3D ) . This reduction in esterification rate was not seen in the progenitor strains Δtether or ice2Δ and could be restored to WT levels by expressing the ER-PM staple , or by growing the cells in choline ( Fig 3D , S7A and S7B Fig ) . The latter result ( i ) suggests that sterol transport between the PM and ER does not depend on ER-PM MCSs , as these structures are equally absent in Δ-s-tether cells grown with or without choline ( Fig 2F ) , and ( ii ) argues against a recent proposal [54] that the sterol acyl transferases Are1 and Are2 act in trans at ER-PM MCSs , directly receiving sterols from the ATP-binding cassette ( ABC ) transporters Aus1 and Pdr11 , thereby eliminating the need for STP-mediated sterol transport between the PM and ER . Transport-coupled esterification of DHE is a complex process that can be separated into a series of discrete mechanistic steps ( Fig 3A ) : ( 1 ) insertion of DHE into the PM , requiring the ABC transporters Aus1 and Pdr11; ( 2 ) equilibration of DHE amongst PM sterol pools , e . g . , pools located in the outer and inner leaflets; ( 3 ) non-vesicular transport of DHE from the cytoplasmic face of the PM to the ER ( 3a ) , a process that requires resumption of ergosterol synthesis ( 3b ) as the cells recover from hypoxia , and transport of ergosterol to the PM ( 3c ) ; and , finally , ( 4 ) esterification of DHE at the ER by the acetyl-CoA acyltransferase ( ACAT ) enzymes Are1 and Are2 . Defects in one or more of these steps could account for the slowdown in DHE esterification seen in Δ-s-tether cells . We verified that DHE loading ( step 1 ) ( Fig 3E ) and ACAT activity ( step 4 ) ( Fig 3F ) were similar in WT and Δ-s-tether cells grown in the absence of choline , and the same was true when the cells were grown in the presence of choline ( S7 Fig [panels C and D] ) . However , the level of endogenous ergosterol in Δ-s-tether cells at the start of the aerobic chase was higher than in WT cells on a per cell basis , although it reached the same value at the end of the chase , indicating that ergosterol resynthesis ( step 3b ) occurs normally ( Fig 3G ) . No difference between ergosterol content and resynthesis was seen when the cells were grown in the presence of choline ( S7E Fig ) . As ergosterol synthesis is largely abolished under hypoxic conditions , the ergosterol content of each cell diminishes with each cell division and is replaced in our protocol by DHE . Because Δ-s-tether cells grow slowly in the absence of choline , the ergosterol “wash-out” is less complete for these cells than for WT cells . The presence of a significant amount of residual ergosterol in Δ-s-tether cells at the start of the aerobic chase could conceivably reduce the rate at which newly synthesized ergosterol is able to displace DHE from the PM , resulting in an apparently slower DHE esterification rate and obscuring information on whether sterol transport between the PM and ER is indeed affected . Thus , our results suggest that the slow rate of esterification observed in Δ-s-tether cells could be due to a defect in steps 2 ( DHE equilibration within the PM ) and/or 3 ( sterol [DHE and ergosterol] exchange between the PM and the ER ) ( Fig 3A ) . To test directly whether sterol exchange between the ER and PM is affected in Δ-s-tether cells , one of the possibilities suggested by our results on retrograde sterol transport ( Fig 3C and 3D ) , we used a pulse-chase protocol ( Fig 4A ) to compare the rate at which newly synthesized ergosterol is transported from the ER to the PM [14 , 21] . The assay was performed as described previously , using [3H]methyl-methionine to pulse-radiolabel ergosterol in the ER [21] . Aliquots of cells taken at different chase time points were homogenized , and the PM was separated from the ER and other internal membranes by sucrose gradient centrifugation . For each time point , the specific radioactivity of [3H]ergosterol ( SR = scintillation counts [cpm] ÷ absorbance at 280 nm ) was determined for the unfractionated cell homogenate and specific fractions after resolving the corresponding lipid extracts by HPLC , and the relative specific radioactivity for each fraction ( RSRfraction = SRfraction ÷ SRcell ) was calculated . Identical subcellular fractionation profiles were obtained with WT and Δ-s-tether homogenates ( Fig 4B ) , displaying clear separation of the PM from internal membranes , as judged by immunoblotting using antibodies against organelle-specific proteins . The quality of the fractionation was exactly as reported in a previous study , in which a wide spectrum of antibodies was used to confirm the separation of the PM from other membranes [21] . The majority of ergosterol was recovered in the PM fraction from WT cells , as expected , and this was also the case for Δ-s-tether cells , indicating that the subcellular distribution of ergosterol is not affected by the absence of ER-PM contact sites ( Fig 4B , bottom panel ) . We analyzed fractions 7 ( PM ) and 2 ( ER-enriched; we designated this fraction ER* to indicate that it contains other intracellular membranes [Fig 4B] ) . The results are shown in Fig 4C . For both WT and Δ-s-tether cells , RSR for ER* was high ( >2 . 0 ) on completion of the labeling pulse because [3H]ergosterol is synthesized in the ER before declining over the chase period to reach a value of 1 . 0 . Conversely , RSR for the PM started at a low level ( <0 . 5; the nonzero value indicates that [3H]ergosterol is transported to the PM even as it is being synthesized during the pulse-labeling period ) and increased to 1 . 0 by the end of the chase . The final RSR values of 1 . 0 for both fractions indicate equilibration of the ergosterol pulse between the ER and PM , as previously reported [14 , 21 , 55] . Mono-exponential fits of the data indicate that [3H]ergosterol is exchanged between the ER and PM with a half time of about 10 min for both WT and Δ-s-tether cells . Thus , the exchange of newly synthesized ergosterol between the ER and PM is normal in Δ-s-tether cells . We considered the possibility that conventional vesicular transport might deliver sterols to the PM to compensate for the possible failure of non-vesicular modes of transport in Δ-s-tether cells . To test this possibility , we constructed a Δ-s-tether sec18-1ts strain that eliminates both exocytosis and ER-PM contact at elevated temperatures . Sec18 is required for exocytosis and most modes of vesicular trafficking [56 , 57] , and the sec18-1ts conditional mutation blocks vesicular transport of secretory proteins and lipids to the PM [56–58] . Whether on its own or in the context of the Δ-s-tether mutations , the sec18-1ts allele does not allow cells to grow at 37 °C , and Δ-s-tether sec18-1 cells do not even grow at 30 °C ( S8 Fig ) . However , the combined growth defects of the Δ-s-tether mutations and sec18-1 at 30 °C are additive , as would be predicted for unrelated pathways , and do not correspond to a synergistic interaction , as would have been observed between mutations disrupting convergent pathways . After culturing at 23 °C , strains were incubated for 20 min at 37 °C and pulse-labeled with [3H]methyl-methionine followed by a 15 min chase . The calculated RSRs of [3H]ergosterol in PM fractions showed no significant differences between WT , sec18-1 , Δ-s-tether , and Δ-s-tether sec18-1 cells , indicating that ergosterol exchange between the ER and PM is unaffected in all of these strains ( Fig 4D ) . These results indicate that secretory vesicles do not provide a compensatory sterol transport mechanism in Δ-s-tether cells . We conclude that the exchange of newly synthesized ergosterol between the ER and PM does not require ER-PM contact sites . This result has two clear implications . First , the sterol transfer machinery in yeast is either absent from or not uniquely localized to ER-PM MCSs . This suggests that contact site–localized proteins such as Lam1–Lam4 are not essential for sterol exchange between the PM and ER; these proteins may act redundantly with soluble STPs or play other roles in intracellular sterol homeostasis [30] . Second , yeast cells do not possess soluble STPs capable of lowering the energy barrier for sterol desorption by >10 kBT , which would make intracellular sterol transport diffusion-limited rather than desorption-limited [12] . Thus , non-vesicular sterol transport in yeast is likely mediated by cytoplasmic STPs that lower the energy barrier for desorption by a more typical 2–3 kBT [12] and that are present in a sufficient number per cell to account for the measured sterol exchange rate [12] . Osh4 is one subset of Osh proteins capable of binding sterols , and it is present in high levels in yeast at >30 , 000 copies per cell [59] . Although elimination of Osh4 had no effect on sterol transport , as measured via assays of sterol import [20] or ER-PM sterol exchange ( S9 Fig ) , we tested whether Osh4 might nevertheless provide a compensatory sterol transport mechanism to allow normal ER-PM sterol exchange in Δ-s-tether cells , where the absence of ER-PM contact sites would prevent any putative membrane-bound STPs from reaching their target membrane . Consistent with a possible redundancy between Osh4 and ER-PM contact sites in sterol transport , we discovered that osh4Δ Δtether cells grew poorly and osh4Δ Δ-s-tether cells were inviable ( Fig 5A ) . In contrast , deletion of the putative sterol transporter encoded by LAM2 had no impact on Δ-s-tether cells , whether cultured with or without added choline ( S10 Fig ) . This result is consistent with the fact that the Δ-s-tether mutations compromise Lam2 function by eliminating its proximity to the cell cortex ( S2 Fig ) , and therefore no further effect would be anticipated on eliminating expression of the protein itself . Expression of Scs2 from a plasmid rescued osh4Δ Δ-s-tether cell lethality , but choline supplementation did not ( Fig 5A ) . This result suggests that an ER-PM tether is required to restore viability to these strains; tether-independent , choline-induced phospholipid synthesis is either irrelevant to osh4Δ Δ-s-tether synthetic lethality , or choline supplementation is simply insufficient to overcome the severity of the phospholipid defect . To test if osh4Δ Δ-s-tether synthetic lethality results from defects in intracellular sterol transport , we generated a conditionally viable strain that combines Δ-s-tether mutations with a temperature-sensitive osh4-1 allele [60] . At 36 °C , osh4-1 osh4Δ Δ-s-tether cells do not grow , and so we measured DHE transport from the PM to the ER 1 h after switching to the inactivating temperature . When OSH4 is inactivated in Δ-s-tether cells in this manner , DHE transfer and esterification were found to be the same as in Δ-s-tether cells ( S11 Fig ) . We conclude that the elimination of OSH4 has no further impact on sterol transport from the PM to the ER in Δ-s-tether cells . Each of the seven OSH genes can provide the essential requirement for the entire family of OSH genes [22] . Even though they are defined as “OSBP homologues , ” not all Osh proteins are able to bind sterols , but all likely bind PI4P [61 , 62] . Thus , Osh6 binds PI4P and PS in a mutually exclusive fashion but cannot bind sterols [23 , 24] . As another way to determine if the osh4Δ Δ-s-tether synthetic lethality relates to the sterol- versus PI4P-binding activities of Osh4 , we therefore tested if Osh6 could functionally replace Osh4 in this context . As shown in Fig 5B , expression of Osh6 from a multicopy plasmid rescued the growth defect of osh4Δ Δ-s-tether cells . Plasmid-based expression of Osh6 was important for growth rescue , as the chromosomally expressed protein , present at fewer than 2 , 000 copies per cell [59] , was not able to support growth . Lipidomics analysis of osh4Δ Δ-s-tether cells rescued with multicopy OSH6 did not reveal an obvious mode of suppression by changes in lipid metabolism ( S6C Fig ) . Compared to Δ-s-tether cells , levels of most sphingolipid precursors and phospholipids ( including PS ) were unchanged in the OSH6-rescued cells or showed minor reductions . The minor reduction in free ergosterol measured in Δ-s-tether cells was restored to WT levels in OSH6-rescued osh4Δ Δ-s-tether cells , and ergosterol ester levels doubled over WT . These results are consistent with a model in which Osh4 and ER-PM tethers function redundantly and independently , but in an important function revolving around PI4P , with indirect effects on sterol metabolism . To explore this model further , we used the PI4P marker GFP-PHOsh2 to compare the distribution of PI4P in WT and Δ-s-tether cells . It had been previously reported that PI4P was dysregulated in Δtether cells [5] and we anticipated that this phenotype might be exacerbated in Δ-s-tether cells . WT cells showed PI4P concentrated at the PM only in buds and also localized to the Golgi apparatus ( Fig 5C and 5D ) ; this distribution was disrupted in Δ-s-tether cells , in which PI4P was evenly distributed throughout the PM in both mother cells and buds ( Fig 5C and 5D ) . The intensity of PI4P staining in the PM of Δ-s-tether mother cells was greater than that seen for Δtether cells ( Fig 5C and 5D ) . To test if the artificial staple could correct the PI4P accumulation/depolarization phenotype , GFP-PHOsh2 and the ER-PM staple were both expressed in Δ-s-tether cells ( S12 Fig ) . Although at a gross level , the artificial tether did not restore normal PI4P polarization , PI4P was absent at the immediate cortical sites where the staple interacted with the PM , suggesting a potentially local corrective effect . The addition of choline to Δ-s-tether cells had no impact on GFP-PHOsh2 depolarization ( 100% of Δ-s-tether cells cultured with or without choline had equal GFP-PHOsh2 fluorescence in mother and bud PM , compared to 4 . 7% of WT cells grown with no added choline and 3 . 6% of WT cells with choline; n > 104 cells ) , indicating that ER-PM MCS regulation of PI4P in the PM is distinct from the role of MCSs in PC metabolism . Taken together , our results suggest that the synthetic lethality of the Δ-s-tether mutations with osh4Δ is associated with dysregulation of PI4P homeostasis . Osh4 and several other Osh proteins have been shown to be upstream regulators of the PI4P phosphatase Sac1 , inducing its activity and thereby affecting PI4P levels in several cellular membranes , including the PM [63] . Sac1 is an ER-membrane protein that interacts with most of the ER-PM tethers deleted in the Δ-s-tether strain [5] , placing it in a position to act across ER-PM contact sites to dephosphorylate PM-localized PI4P . If Osh4 acts through Sac1 in the same pathway , then sac1Δ might also be synthetically lethal in Δ-s-tether cells . This was indeed the case ( Fig 5E ) , consistent with the model that Osh4 and Sac1 function in a PI4P regulatory pathway operating alongside ER-PM tethers . One possibility is that Sac1 itself provides limited tethering , a function that might be induced by the absence of the other tethers in Δ-s-tether cells . However , expressing the soluble enzymatic domain of Sac1 ( Sac11–522 ) without its ER membrane-binding domain suppressed sac1Δ Δ-s-tether lethality ( S13A Fig ) . This result indicated that Sac1 does not act as a tether and that Sac1 dependence on MCSs can be partially bypassed if Sac1 is released from the membrane , so that it can access PI4P in the PM . Because Osh4 acts in vitro as a soluble PI4P transport protein [64] , it might function in cells to extract and transport PI4P from the PM to Sac1 in the ER . We tested if the requirement for Osh4-mediated PI4P transport could be circumvented if Sac1 was freed from the ER membrane to diffuse to the PM to dephosphorylate PI4P . However , soluble Sac11–522 expressed from a multicopy plasmid did not rescue osh4Δ Δ-s-tether lethality , indicating that the requirement for Osh4 cannot be bypassed by liberating Sac1 from the ER ( S13B Fig ) . Thus , in the context of ER-PM MCSs , Osh4 might play an important role in Sac1 regulation , but it clearly has other independent functions as well . We have shown that the absence of ER-PM contact sites does not affect sterol exchange between the ER and PM ( Fig 4C ) and that this lack of effect is not due to compensatory sterol transport by secretory vesicles ( Fig 4D ) or by the single most abundant sterol-binding protein in yeast , Osh4 ( Fig 5B ) . We can therefore pinpoint the cause of slow transport-coupled esterification of exogenously supplied DHE in Δ-s-tether cells to step 2 in the scheme depicted in Fig 3A , i . e . , the exchange of sterol between sterol pools within the PM lipid bilayer . To investigate this point , we tested the growth of Δ-s-tether cells in the presence of three drugs that report on the lipid organization of the PM: nystatin , duramycin , and edelfosine . Nystatin is an ergosterol-binding polyene antimycotic compound . Nystatin resistance is observed in viable sterol biosynthesis mutants and some mutants , such as osh4Δ [51] , that disrupt sterol organization within the PM . Conversely , many mutants with altered lipid composition and/or PM organization exhibit nystatin sensitivity [51 , 65] . On nystatin-containing medium , Δ-s-tether cells exhibited an exacerbated growth sensitivity compared to Δtether cells , and both strains were more sensitive than WT or nystatin-resistant osh4Δ cells ( Fig 6A ) . Duramycin is a lantibiotic that disrupts cell growth by directly binding PE in the outer leaflet of the PM . As PE is principally located in the cytoplasmic leaflet of the PM , duramycin sensitivity indicates changes in PE bilayer asymmetry , as seen in the phospholipid-flippase mutant lem3Δ . Growth of WT , Δtether , and Δ-s-tether cells was not significantly affected by duramycin ( Fig 6B ) , indicating that transbilayer phospholipid asymmetry is unaffected . We next tested edelfosine , a cytotoxic lysophosphatidylcholine analogue whose activity in yeast is modulated by PM phospholipid flippase activity , and by sterol and sphingolipid pathways . A flippase defect confers edelfosine resistance [66] , whereas changes in the lipid composition and physical properties of the PM confer edelfosine sensitivity [67] . Δ-s-tether cells displayed acute cytotoxicity to edelfosine compared to WT or even Δtether cells ( Fig 6B ) , consistent with changes in PM properties . Based on the sensitivity of Δ-s-tether cells to edelfosine ( Fig 6B ) as well as the significant reductions in their sphingolipid levels revealed by lipidomics analyses ( Fig 2G ) , we considered the possibility that the cells would exhibit a growth phenotype in response to the sphingolipid synthesis inhibitor myriocin ( Fig 6C ) . Indeed , previous work had shown that elimination of the three Tcbs alone causes myriocin sensitivity [35] . Unexpectedly , both Δtether and Δ-s-tether cells were myriocin resistant ( Fig 6C ) . These results suggest that myriocin toxicity in Δ-s-tether cells is mitigated by compensatory alterations either in membrane composition or in the sphingolipid biosynthesis apparatus . Taken together , the results of our drug screening experiments indicate that changes in PC and sphingolipid organization in Δ-s-tether cells might indirectly modulate sterol pools within the PM . The perturbation in PM lipid organization revealed by drug tests ( Fig 6A–6C ) was not evident in measurements of the ergosterol “status” of the cell ( S14 Fig ) . Thus , when comparing WT and Δ-s-tether cells , we found no significant difference in the fraction of total cellular ergosterol that was recovered in detergent-insoluble membranes ( DIMs ) ( S14F Fig ) , a crude readout of the extent to which ergosterol associates with phospholipids and sphingolipids containing saturated acyl chains [14 , 21] . Likewise , there were no significant differences in the total ergosterol content of the cells ( S14A Fig ) , the ergosterol/phospholipid ratio ( S14B Fig ) , or the fraction of cellular ergosterol located in the PM ( Fig 4B ) . Because these bulk measurements are unlikely to be responsive to nuanced changes in lipid composition and organization , we chose a more sensitive technique to probe ergosterol organization at the PM . Methyl-β-cyclodextrin ( MβCD ) extracts only a very small fraction , <0 . 5% , of total cellular ergosterol from the outer leaflet of the PM of WT cells under our standard conditions [14 , 21] , indicative of the unusual physical properties of the yeast PM [14 , 21 , 68–70] . When PM lipid organization is perturbed , then the amount of MβCD-extractable sterol can increase dramatically , as seen as in oshΔ osh4-1ts cells and sphingolipid-deficient lcb1-100ts cells at the nonpermissive temperature [14 , 21] . We compared the MβCD-extractability of ergosterol in WT cells versus the tether mutants ( Fig 6D ) . As reported previously , the proportion of ergosterol extracted from WT cells by MβCD is about 0 . 25% of total cellular ergosterol [14 , 21]; a similarly low level of extraction ( <1% ) was obtained with Δtether cells ( Fig 6E ) . However , in Δ-s-tether cells , the MβCD-accessible ergosterol pool in the PM was >5% , about 20-fold greater than for WT cells , consistent with a major change in the PM lipid bilayer that enabled greater extraction of ergosterol . This effect was largely reset by expression of the ER-PM staple and completely restored to WT levels by supplementing the growth medium with choline ( Fig 6E ) . The ability of both the ER-PM staple and choline to restore PM lipid organization , as revealed by MβCD-extractability of ergosterol , parallels their ability to correct the slowdown in retrograde transport of DHE ( Fig 3D ) . Thus , these results are consistent with the idea that the reduced rate of transport-coupled esterification of DHE is due to perturbations of the PM lipid bilayer that delay the access of exogenously supplied DHE to cytoplasmic STPs ( Fig 3D , step 2 ) . The ability of choline to provide the same corrective effect as the ER-PM staple without inducing membrane contacts indicates that the role of tethers in this context is to support normal phospholipid and/or sphingolipid homeostasis , and thereby membrane organization . Although the exchange of ergosterol between the ER and PM as a whole was unchanged in Δ-s-tether cells ( Fig 4C and 4D ) , we investigated if movement of ergosterol within the PM bilayer might be affected . Non-vesicular transport of newly synthesized [3H]ergosterol deposits ergosterol molecules in the cytoplasmic leaflet of the PM . At a minimum , these molecules must exchange with the outer leaflet pool of ergosterol before they fully equilibrate with PM ergosterol pools and become accessible to MβCD ( Fig 6F ) . We tested if the exchange of ergosterol within the PM was affected in Δ-s-tether cells by measuring the rate at which newly synthesized ergosterol becomes accessible to MβCD extraction . We used [3H]methyl-methionine to pulse-label ergosterol in the ER and then chased the cells for 30 min . The samples were subjected to MβCD extraction and , in parallel , samples were taken for subcellular fractionation to isolate the PM ( as in Fig 4B ) . Both the MβCD extract and the PM fraction were processed with organic solvents to extract ergosterol for HPLC analysis and measurement of RSR . The RSR for the PM fraction after a 30 min chase was about 0 . 8 for both WT and Δ-s-tether cells ( Fig 6G , dashed line ) , as expected ( Fig 4C ) . However , the RSR for MβCD-extracted ergosterol in WT cells was about 0 . 65 ( Fig 6G , WT ) , indicating a slight delay in the transport of ergosterol within the PM to the MβCD-accessible pool in the outer leaflet , consistent with our previous report [21] . This delay was considerably greater in Δ-s-tether cells , where the MβCD-extracted ergosterol had an RSR of only about 0 . 15 after a 30 min chase ( Fig 6G , Δ-s-tether ) . Expression of the ER-PM staple reduced the delay significantly , such that the RSR increased to about 0 . 3 in Δ-s-tether cells chased for 30 min ( Fig 6G , Δ-s-tether + staple ) . We conclude that ( i ) the transfer of ergosterol from its site of arrival at the cytoplasmic leaflet of the PM to the outer leaflet pool , from which it can be extracted by MβCD , is slower than the rate at which ergosterol exchanges between the ER and PM as a whole , as reported previously [21] , and ( ii ) the intra-PM movement of ergosterol , from the inner to the outer leaflet , is dramatically slower in Δ-s-tether cells compared with WT cells . Taken together with the fact that the abundance of characteristic PM lipids , e . g . , IPC , MIPC , and PS , in Δ-s-tether cells differs significantly from WT cells ( Fig 2G ) , it seems likely that the changes in ergosterol organization in the PM and the rate of exchange between ergosterol pools in the PM are an indirect consequence of changes in PM phospholipid and sphingolipid composition . We have shown that membrane contact between the ER and PM impacts the abundance of PM lipids ( sphingolipids , PE , PS [Fig 2G] , and PI4P [Fig 5C] ) , PM lipid organization ( Fig 6A , 6B , 6C and 6E ) , and the intra-PM movement of ergosterol ( Fig 6G ) . In turn , it is known that PM lipids play a role in the establishment of contact sites ( Fig 1A ) ; thus , phosphoinositides and PS in the PM provide anchors for ER-localized Tcb1–Tcb3 , Ist2 , and Scs2 [6 , 8 , 36–40] . As sterols represent a large fraction of PM lipids and are critical determinants of PM organization [10 , 11] , we analyzed the potential role of sterols in establishing contact sites between the ER and PM . To test the dependence of MCS formation on ergosterol , we depleted yeast cells of sterols and visualized cER-PM association by both transmission electron microscopy and Tcb3-GFP and RFP-ER distribution by fluorescence microscopy . Squalene synthase ( Erg9 ) represents the first sterol-specific enzymatic step in the production of all sterols , and inhibition of Erg9 specifically blocks sterol synthesis without directly affecting other isoprenoids [71] . In erg9Δ PMET3-ERG9 cells , methionine addition to the growth medium represses Erg9 expression and de novo sterol synthesis stops . To our surprise , electron microscopy showed that sterol depletion in erg9Δ PMET3-ERG9 cells resulted in a dramatic expansion of cER ( Fig 7A ) , such that the inner face of the PM was nearly completely covered with associated ER membrane ( Fig 7B ) . This finding was confirmed by confocal fluorescence microscopy in live Tcb3-GFP–expressing cells . In sterol-replete WT cells , Tcb3-GFP fluorescence exhibited a characteristic discontinuous stitched pattern around the cortex ( Fig 7C ) [35] . In about 90% of sterol-depleted erg9Δ PMET3-ERG9 cells , however , cortical fluorescence was essentially contiguous ( Fig 7C ) . Although sterol-depleted cells accumulate as unbudded cells in the G1-phase of the cell-cycle , G1-arrested cdc42-101 cells did not induce any change in Tcb3-GFP distribution , indicating that increased ER-PM contact is not due to G1 arrest per se ( S15 Fig ) . These results indicate that ER-PM membrane association is induced when cellular sterol synthesis is blocked . In addition to its altered distribution along the PM , Tcb3-GFP fluorescence was generally greater in sterol-depleted cells relative to WT , suggesting an induction of Tcb3 protein levels in response to sterol reduction . This point was verified by analyzing cell extracts prepared from both WT and erg9Δ PMET3-ERG9 sterol-depleted cells expressing Tcb3-GFP and determining relative levels of Tcb3-GFP by SDS-PAGE/immunoblotting using anti-GFP antibodies . When normalized to levels of the actin ( Act1 ) internal control , Tcb3-GFP protein levels were seen to be induced about 6-fold in sterol-depleted cells , compared to similarly treated WT cells ( Fig 7D ) . In genome-wide analyses of gene expression by DNA microarray , sterol depletion had no impact on transcript levels of any of the tether genes; relative to WT cells , methionine repression of de novo sterol synthesis in erg9Δ PMET3-ERG9 cells showed transcriptional changes between 0 . 93 and 1 . 05 ± 0 . 03 ( mean ± SD; independent duplicate trials ) for each of the seven tether protein genes . These results indicated that Tcb3 protein levels are posttranscriptionally regulated . Because of the long half-life of cellular sterols , an extended period is required after ERG9 repression for complete sterol depletion . To determine how quickly sterol-depleted cells recover their normal distribution of ER-PM association , ER-RFP and Tcb3-GFP redistribution was measured in response to exogenously added cholesterol . Under standard culture conditions , yeast does not import sterols from the medium as discussed above ( Fig 3 ) , but the deletion of HEM1 permits cholesterol uptake [72] . A hem1Δ erg9Δ PMET3-ERG9 strain could grow after sterol depletion , but only when exogenous cholesterol ( or δ-aminolevulinic acid [δ-ALA] , the product of the Hem1 enzyme ) was supplemented to the growth medium ( S16 Fig ) . In hem1Δ cells , ER-RFP and Tcb3-GFP distributions were the same with or without cholesterol supplementation ( Fig 7E ) . In sterol-depleted hem1Δ erg9Δ PMET3-ERG9 cells , return to the normal discontinuous stitched fluorescence of cortical Tcb3-GFP commenced 1 h after cholesterol addition , and the characteristic WT pattern was observed after about 4 h ( Fig 7E and 7F ) . In these cells , recovery of normal ER-RFP morphology lagged behind the restoration of the normal Tcb3-GFP distribution ( Fig 7F ) , consistent with the idea that tethering complexes dictate changes in cER association . These results indicate that tethering between the ER and PM responds to cellular sterol pools .
To create Δ-s-tether cells , we eliminated ICE2 in the previously described Δtether strain . The role of Ice2 in distributing ER along the PM between mother and daughter cells during mitosis is well established [33 , 44] , hinting that it may play a direct role in tethering ER to the PM . Ice2 is a polytopic ER membrane protein with a single prominent cytoplasmic loop that has been implicated in associating the ER with lipid droplets during the stationary phase of growth , and potentially channeling DAG to the phospholipid biosynthetic machinery in the ER as cells resume growth [45] . In analogy to its proposed tethering role in stationary phase cells , we speculate that Ice2 may play a role in bridging the ER and PM in rapidly dividing cells . Indeed , fluorescence microscopy reveals that Ice2 is located at the cell cortex in Δtether cells ( S3 Fig ) . As a potential tether protein , the cytoplasmic loop of Ice2 may interact directly in trans with the cytosolic face of the PM or , similar to the Scs2 tether [5 , 73] , Ice2 might form a bridge across the ER-PM interface via an interaction with another protein . If the latter scenario is correct , then the mechanism of tethering by both Ice2 and Scs2 would differ from that of the autonomous membrane attachments conferred by Ist2 and the E-Syt homologues Tcb1–Tcb3 ( Fig 1A ) . Nevertheless , eliminating Ice2 in the context of Δtether cells results in quantifiable reductions in ER-PM association beyond those previously reported for Δtether cells ( Fig 1E and S1 Fig ) , leading to clear functional outcomes . For example , synthetic lethality of Δ-s-tether with osh4Δ or sac1Δ was not manifested in the progenitor Δtether strain and only occurred with the additional deletion of ICE2 . Likewise , slowing of transport-coupled esterification of DHE ( Fig 3C ) and increased extractability of ergosterol by MβCD ( Fig 6E ) were observed only after deletion of ICE2 in Δtether cells . Taken together , these findings show that Ice2 is an important contributor to ER-PM tethering and associated functions . We found that bidirectional sterol exchange between the ER and PM occurs at the same rate in Δ-s-tether and WT cells , indicating that ER-PM contact sites do not contribute quantitatively to the mechanism of sterol movement between these two membranes . If any elements of the sterol transport machinery are localized to ER-PM MCSs , then their function must be subsumed by other sterol transport mechanisms in Δ-s-tether cells . However , we show that in and of themselves , neither secretory vesicles ( Fig 4D ) nor the cytoplasmic sterol-binding protein Osh4 ( Fig 5A and 5B ) or the ER-anchored Lam2 protein ( S10 Fig ) provide this putative compensatory mechanism in Δ-s-tether cells . Our results also make it clear that yeast cells do not possess STPs with the ability to lower the energy barrier for sterol desorption to the point at which transport becomes a diffusion-limited rather than desorption-limited process [12] . Thus , non-vesicular sterol transport in yeast is likely mediated by unremarkable cytoplasmic STPs ( i . e . , STPs that are able to lower the energy barrier for sterol desorption by only 2–3 kBT ) present in a sufficient number per cell to account for the measured sterol exchange rate [12] . The identification of these STPs is a focus of future work . Even though ER-PM sterol exchange was unaffected by the lack of ER-PM contact sites , trafficking of exogenously supplied DHE to the ER was unexpectedly slow in Δ-s-tether cells compared with WT cells ( Fig 3C ) . Careful analysis of the various mechanistic steps of the transport process ( Fig 3A ) revealed that the slowdown could be linked to a dramatic lowering of the rate at which sterols equilibrate between the inner and outer leaflet of the PM in Δ-s-tether cells ( Fig 6F and 6G ) . While this slowdown should not affect the equilibration of DHE with PM sterol pools during the extended hypoxic loading period used for this assay ( Fig 3A , steps 1 and 2 ) , it would affect the rate at which newly synthesized ergosterol displaces DHE from the PM during the aerobic chase , thereby resulting in a slower rate of transport-coupled DHE esterification . We previously reported that the appearance of newly synthesized ergosterol in the MβCD-extractable ergosterol pool in the outer leaflet of the PM lags behind its arrival at the PM in WT cells ( Fig 6D and 6E ) , suggesting that equilibration of sterol across the yeast PM is considerably slower than that seen for cholesterol flip-flop in synthetic , liquid crystalline membranes and in red blood cells [74–76] . This may be a consequence of the unusual properties of the yeast PM , exemplified by the slow lateral diffusion of both lipids and proteins [69 , 70] and the organization of PM proteins into a mosaic of domains [77] . In the case of Δ-s-tether cells , the rate of transbilayer sterol equilibration was about 5-fold slower than for WT cells ( Fig 6G ) , and this was reflected in changes in PM bilayer organization , as manifested in the nystatin and edelfosine sensitivity of these cells and the greater accessibility of sterols to MβCD extraction ( Fig 6A , 6B and 6E ) . Quantification of cellular lipids revealed reductions in PE , PS , and the sphingolipids IPC and MIPC ( Fig 2G ) . As these lipids generally reflect PM composition [78] , we propose that relative changes in lipid levels are the underlying cause of the disturbance in the PM bilayer , resulting in a change in ergosterol dynamics . How do ER-PM contact sites affect PM lipid organization and intra-PM ergosterol dynamics ? Because the growth defect of Δ-s-tether cells was rescued by choline supplementation ( Fig 2E ) , which increases flux through phospholipid biosynthetic pathways [79] without inducing contact site formation ( Fig 2F ) , the primary defect in cells lacking ER-PM contact sites appears to involve phospholipid regulation . Remarkably , both choline and the artificial staple corrected most defects inherent to Δ-s-tether cells ( i . e . , slow growth [Fig 2B] , slow transport-coupled DHE esterification [Fig 3D] , high MβCD-extractability of ergosterol [Fig 6E] , and slow rate of ergosterol exchange between PM pools [Fig 6G] ) . These results indicate that the function of the endogenous tethers , even those such as Tcb1 , Tcb2 , and Tcb3 , with lipid-transporting SMP domains [48] might be largely structural in this context , i . e . , the tethers provide a means of mechanical attachment of the ER to the PM , thereby enabling ER-localized proteins , such as the PC-synthesizing phospholipid methyltransferase Opi3 , to act in trans . An exception to this general conclusion is that neither choline nor the artificial staple was able to re-establish normal PI4P polarization in Δ-s-tether cells ( Fig 5C and 5D and S12 Fig ) . PI4P dephosphorylation at the PM is proposed to be due to the ER-localized Sac1 phosphatase acting in trans at ER-PM contact sites [80] , although the protein clearly also acts in cis [81] . The inability of the artificial staple to facilitate PM access for Sac1 might stem from either ( i ) an insufficient number or improper positioning of membrane contacts established by the staple and/or ( ii ) the inability of the staple to provide a specific requirement for Sac1 activation , which is otherwise provided by endogenous tethers ( almost all tethers were identified by virtue of specific interactions with Sac1 [5] , whereas the artificial tether would lack this ability ) . Our data suggest that PM PI4P is reduced in the immediate vicinity of contact sites generated by the artificial staple ( S12B Fig ) , indicating that Sac1 might access PI4P locally at those points . In contrast , endogenous tethers and their ancillary factors might further expand cER-associated regions of the PM that are accessible to Sac1 . In yeast , all phospholipids are synthesized from phosphatidic acid ( PA ) via the CDP-DAG pathway; PE and PC are also synthesized by the Kennedy pathway using DAG and salvaged or exogenously supplied ethanolamine and choline [79 , 82] ( S17 Fig ) . PA levels and the DAG:PA ratio are critical in determining the amount of CDP-DAG available for phospholipid synthesis ( S17 Fig ) . The mole percentage of PA in Δ-s-tether cells is 20% lower than that in WT cells , and the DAG:PA ratio is 2-fold greater ( Fig 2G ) , indicating dysregulation of phospholipid synthesis in the absence of ER-PM contact sites . Consistent with this , levels of PS and PI-derived sphingolipids are much lower in Δ-s-tether than in WT cells ( Fig 2G ) , although PI levels are unaffected ( see below ) . As previously shown [31] , ice2Δ scs2Δ cells have a diminished ability to convert PS to PC via Opi3-mediated phospholipid methylation at ER-PM contact sites , necessitating choline supplementation for normal growth . In Δ-s-tether cells , choline supplementation would not only bypass the need for Opi3 but also compensate for the lower overall rate of phospholipid synthesis resulting from decreased PA levels by generating PC via the Kennedy pathway for membrane growth . Lipidomic analysis of Δ-s-tether cells cultured with choline did in fact show restoration of PC to levels comparable to WT ( S6A Fig ) . These results are also consistent with the choline-reversible synthetic growth defects observed when either OPI3 or CHO2 is deleted in Δ-s-tether cells ( S4A Fig ) . Thus , ER-PM contact sites may function as regulatory interfaces that coordinate the CDP-DAG and Kennedy pathways to balance convergent mechanisms for phospholipid synthesis . Δ-s-tether cells have a normal mole percentage of PI and increased PI4P , yet their content of inositol sphingolipids is about 60% lower than in WT cells ( Fig 2G ) , indicating dysregulation of phosphoinositide homeostasis because of loss of ER-PM contacts . Total cellular PI is generated predominantly by the CDP-DAG pathway and Sac1-mediated dephosphorylation of PI4P ( itself generated from PI ) in separate cellular locations , with both routes providing the biosynthetic precursor for complex inositol sphingolipids ( S17 Fig ) . Brice and colleagues [83] reported that disruption of SAC1 alone reduces PI levels dramatically and IPC and MIPC levels by >70% . The Δ-s-tether mutations would reduce the Sac1-mediated route for PI production by distancing the enzyme from its substrate in the PM ( Fig 5C ) . However , sac1Δ and Δ-s-tether mutations are lethal when combined , likely due to limiting PI levels , indicating that Δ-s-tether mutations must also inhibit PI production from the CDP-DAG pathway . Normal PI levels in Δ-s-tether cells may therefore be a consequence of preserving this lipid at the expense of inositol sphingolipid production . It should be noted that Sac1 is also a component of the SPOTS ( SPT , Orm1/2 , Tsc3 , and Sac1 ) complex that regulates early steps in sphingolipid synthesis [84 , 85] . However , levels of ceramide , an early precursor in sphingolipid synthesis , were essentially normal in Δ-s-tether cells ( Fig 2G ) . It therefore seems unlikely that the SPOTS complex plays a direct role in ER-PM contact site regulation of inositol sphingolipids . While investigating whether normal ER-PM sterol transport in Δ-s-tether cells could be due to compensatory activity of soluble or membrane-bound STPs , we discovered that the deletion of OSH4 was synthetically lethal with Δ-s-tether mutations ( Fig 5A ) . Lethality was not due to a sterol-related process because Osh6 , which does not bind sterols , could rescue osh4Δ Δ-s-tether growth defects ( Fig 5B ) . Consistent with previous proposals that Osh proteins represent important regulators of PI4P [15 , 80] , the deletion of SAC1 in Δ-s-tether cells also resulted in synthetic lethality ( Fig 5E ) . Based on these findings , we propose that Osh4 ( and Sac1 ) functions in a parallel pathway alongside ER-PM contact sites for PI4P regulation . However , expression of the soluble enzymatic domain of Sac1 did not suppress osh4Δ Δ-s-tether lethality ( S13 Fig ) , suggesting that the downstream regulation of Sac1 is not the only role Osh4 plays at ER-PM MCSs . The availability of Δ-s-tether cells now allows further interrogation of the mechanism by which ER-PM MCSs function as interfaces for regulating PI4P signaling and phospholipid metabolism . Our results are consistent with the hypothesis that ER-PM contact sites constitute a regulatory nexus to balance sterol and phospholipid concentrations to maintain PM structure . In the absence of contact sites , the ratio of sterols to specific phospholipids and sphingolipids is uncoupled , which negatively impacts PM organization , intra-PM sterol dynamics , and PI4P levels . When sterols become limiting , posttranscriptional induction of Tcb3 increases the extent of ER-PM association , potentially facilitating compensatory changes to phospholipid synthesis to re-establish bilayer stability . Although the mechanisms that control Tcb3 levels in sterol-replete or sterol-depleted cells are not known , it is interesting to speculate that the induction of a tether protein may represent a new homeostatic mechanism for regulating PM composition and structure .
Yeast strains and plasmids are listed in the Supporting information ( S1 and S2 Tables , respectively ) . Unless otherwise stated , yeast cultures were grown in synthetic complete or YPD rich media at 30 °C . All temperature-sensitive alleles were cultured at permissive growth temperatures ( 30 °C unless otherwise stated ) and shifted to the restrictive temperature of 37 °C , as specified . DNA cloning and bacterial and yeast transformations were carried out using standard techniques [86 , 87] . For the choline and ethanolamine supplementation growth assays , yeast strains were cultured in synthetic minimal media for 48 h , then streaked onto solid synthetic complete media containing 1 mM choline chloride or 1 mM ethanolamine ( Sigma-Aldrich Chemicals , St . Louis , MO ) . Growth in response to inositol supplementation was tested on synthetic minimal media containing 75 μM myo-inositol ( Sigma-Aldrich Chemicals ) . For the nystatin sensitivity plate assay , 10-fold serial dilutions of yeast cultures were spotted onto solid synthetic media containing 2 . 5 μM nystatin ( Sigma-Aldrich Chemicals ) . Cell growth was also tested on solid rich media containing 60 μM edelfosine ( Cayman Chemical , Ann Arbor , MI ) , 5 μM duramycin ( Sigma-Aldrich Chemicals ) , or 0 . 5 μg/mL myriocin ( Sigma-Aldrich Chemicals ) . To select against URA3-marked plasmids ( e . g . , pCB1183 ) , yeast cultures were grown on rich growth media and then streaked onto a synthetic solid growth medium containing 1 g/L 5-fluoroorotic acid ( Gold Biotechnology , St . Louis , MO ) . For assays of cholesterol uptake by cells lacking HEM1 , yeast cultures were spotted onto a solid synthetic medium lacking methionine , containing 25 μg/mL cholesterol in 1% ( vol/vol ) Tween 80–ethanol ( 1:1 [vol/vol] ) and 50 μg/mL δ-ALA ( Sigma-Aldrich Chemicals ) . For sterol depletion assays , erg9Δ PMET3-ERG9 cells were grown at 30 °C for 10 h in synthetic media lacking methionine and grown to mid-log phase before adding 100 mg/L methionine . DNA cloning and bacterial and yeast transformations were carried out using standard techniques [86 , 87] . The functional artificial tether fusion plasmids pCB1185 and pCB1188 were derived from pRS416-PYSP1-eGFP-Myc-HMH-RitC , a kind gift from Tim Levine ( UCL Institute of Ophthalmology ) . To construct pCB1185 , coding sequences from pRS416-PYSP1-eGFP-Myc-HMH-RitC were amplified using CACTCGAGTTATGGAGCAAAAGCTCATTTCTGAAGAG and CAGGTACCCTATACTGAATCCTTTTTCTTACGGAAT primers , and the product was digested with XhoI/KpnI for subcloning in frame with GFP under the control of an ACT1 promoter in a YCplac111 vector . To construct pCB1188 , coding sequences from pRS416-PYSP1-eGFP-Myc-HMH-RitC were amplified using the primers: CATCCGGACTTATGGAGCAAAAGCTCATTTCTGAAGAG and CATCTAGACTATACTGAATCCTTTTTCTTACGGAATGG . The amplified product was digested with KpnI/XbaI and subcloned in frame with coding sequences for mCherry under the control of an ACT1 promoter in a YCplac111 vector . All genomic manipulations were performed by integration of PCR amplified product as previously described [88] . All natMX4 and hphMX4 deletions were generated by homologous recombination into the yeast genome of targeted P4339 and pAG32 amplified products; transformants were selected for on YPD media containing 100 mg/L nourseothricin ( Gold Biotechnology ) and 400 mg/L hygromycin B ( Toku-E , Bellingham , WA ) , respectively . For growth of hem1Δ::natMX4 cells , selective growth media contained 50 μg/mL δ-ALA . The sec18-1:URA3 temperature-sensitive allele strains were isolated at 23 °C after genomic recombination of the sec18-1:URA3 gene cassette amplified from CBY2853 genomic DNA . All transformants were confirmed by genomic PCR or genetic complementation assays . Yeast cells were grown to mid-log phase and prepared ( fixation , dehydration , infiltration/embedding ) as previously described [89] . Minor changes were made to the infiltration schedule as follows: ethanol:resin ( 2:1 ) was incubated overnight while ethanol:resin ( 1:1 ) was incubated for 5 h . For calculations of cER abundance in electron micrographs , the ratio between PM and the length of PM associated with cER were determined using ImageJ ( www . imagej . nih . gov/ij/index . html ) ; cER was assigned as previously described [90] . The resin block was microtomed to expose a clean face and then attached to a metal SEM stub with carbon tape . The sides were coated with silver paint to increase conductivity . The block was then sputter coated with a thin coat of Au/Pd and inserted into the FIB-SEM . Areas of interest were identified by viewing with the electron beam at 25 keV . Serial block-face imaging: the area of interested was coated with 1 μm-thick Pt in the microscope using the Pt deposition needle . The sample was tilted to 52 degrees and an approximately 30 μm trench was cut in front of the area . The FEI Slice and View G2 program was used for data collection , with the following parameters: imaging at 2 keV; 50 pA current; 30 μs dwell time; horizontal field width , 17 . 74 μm; tilt angle , 60 degrees ( cross-sectional viewing angle , −30 degrees ) ; working distance , 2 . 5 mm; and TLD detector set to −245 V suction tube voltage for backscatter imaging . The slice thickness was set at 20 nm , so the final voxel size was 8 . 66 nm in X , 10 nm in Y , and 20 nm in Z . Image processing: raw images were aligned using the xfalign tool in IMOD [91] . Images were then corrected for density gradients using ImageJ software [92] . The aligned and corrected tiff images were imported into Amira for density-guided segmentation ( FEI Software , Hillsboro , OR ) and display . Confocal fluorescence microscopy was performed as previously described [93] . For all experiments , yeast cells were grown to mid-log phase before visualization . GFP-Staple ( pCB1185 ) and mCherry-Staple ( pCB1188 ) fusion proteins were imaged using 150 and 750 ms exposures , respectively . RFP-ER ( pCB1024 and pCB1277 ) and RFP-RAS2 ( pCB1204 ) were imaged using a 750 ms exposure on the confocal . GFP-2xPHOSH2 ( pTL511 ) was imaged by confocal microscopy using a 250 ms exposure . Tcb3p-GFP and Ice2p-GFP were imaged by confocal microscopy using 350 ms and 1 . 5 s exposures , respectively . Widefield fluorescence microscopy was performed as previously described [94] . GFP-2xPHOSH2 ( pTL511 ) imaged by widefield epifluorescence was acquired using a 200 ms exposure , 30% arc lamp intensity , and analog gain set to full . Bleed-through between fluorescence channels was undetectable under the conditions used for image acquisition . All contrast enhancement was kept constant for each series of images . Transport of exogenously supplied DHE from the PM to the ER was determined as previously described [21 , 51] . Briefly , DHE was loaded into the PM of cells under hypoxic conditions , and its transport to the ER upon subsequent aerobic chase was monitored by fluorescence microscopy and quantified by lipid extraction and HPLC to determine the extent of conversion to DHE-ester . To quantify the initial fluorescence of DHE-loaded cells , individual cells were outlined using ImageJ; then , the corresponding area and integrated density were measured to determine the corrected total cell fluorescence ( CTCF ) , as previously described [95]; CTCF = ( integrated density − ( area of selected cell × mean background fluorescence ) ) . At least 40 cells were counted ( from 4 individual fields ) for each strain . The ACAT activity of the cells was assayed with microsomes using a modification of a published procedure [96] , as described [21 , 51] . Biosynthetic sterol transport was measured using a pulse-chase labeling procedure as previously described [21 , 51] . Briefly , cells were labeled with [3H]methyl-methionine for 4 min to generate a pulse of [3H]ergosterol in the ER and subsequently chased with unlabeled methionine . Transport was assessed after subcellular fractionation to isolate the PM or after MβCD extraction to sample ergosterol in the outer leaflet of the PM . Ergosterol in cells , subcellular fractions , and MβCD extracts were solubilized using organic solvents and quantified by HPLC . For lipidomics analysis , cells were grown to about OD600 0 . 8 and lipids were extracted with chloroform:methanol ( 2:1 ) . Yeast lipid extracts were prepared using a standard chloroform-methanol mixture , spiked with appropriate internal standards , and analyzed using a 6490 Triple Quadrupole LC/MS system ( Agilent Technologies , Santa Clara , CA ) [97] . Glycerophospholipids and sphingolipids were separated with normal-phase HPLC as described before [97] , with a few changes . An Agilent Zorbax Rx-Sil column ( inner diameter 2 . 1 × 100 mm ) was used under the following conditions: mobile phase A ( chloroform:methanol:1 M ammonium hydroxide , 89 . 9:10:0 . 1 , v/v ) and mobile phase B ( chloroform:methanol:water:ammonium hydroxide , 55:39 . 9:5:0 . 1 , v/v ) ; 95% A for 2 min , linear gradient to 30% A over 18 min and held for 3 min , and linear gradient to 95% A over 2 min and held for 6 min . Sterols and glycerolipids were separated with reverse-phase HPLC using an isocratic mobile phase as before [97] , except with an Agilent Zorbax Eclipse XDB-C18 column ( 4 . 6 × 100 mm ) . Quantification of lipid species was accomplished using multiple reaction monitoring ( MRM ) transitions [97 , 98] in conjunction with the referencing of appropriate internal standards: PA 17:0/14:1 , PC 17:0/20:4 , PE 17:0/14:1 , PG 17:0/20:4 , PI 17:0/20:4 , PS 17:0/14:1 , LPC 17:0 , LPE 14:0 , Cer d18:1/17:0 , D7-cholesterol , cholesteryl ester ( CE ) 17:0 , 4ME 16:0 diether DG , D5-TG 16:0/18:0/16:0 ( Avanti Polar Lipids , Alabaster , AL ) . Quality and batch controls [99] were included to assess instrument stability and reproducibility and allow for correction of drift and other systematic noise , e . g . , biases correlated with analysis order and/or sample preparation . Values are represented as mole fraction with respect to total lipid ( mole percentage ) [97] . All lipid species and subclasses were analyzed with one-way ANOVA followed by a post hoc Bonferroni test . For analysis of Tcb3 protein expression , 10 OD600 units of Tcb3p-GFP–expressing cells post sterol depletion were prepared as described by Ohashi and colleagues [100] . Pellets were resuspended in SDS sample buffer and boiled for 5 min before SDS-PAGE . Protein transfer to nitrocellulose membranes and immunoblot conditions were as previously described [101] . To detect Tcb3p-GFP , immunoblots were incubated with a 1:1 , 000 anti-GFP antibody ( ThermoFisher Scientific Inc . , Waltham , MA ) followed with 1:10 , 000 anti-rabbit-HRP secondary antibody ( Bio-Rad Laboratories , Mississauga , ON ) . Actin was detected using 1:1 , 000 anti-actin antibody ( Cedarlane , Burlington , ON ) followed with 1:10 , 000 anti-mouse-HRP secondary antibody ( ThermoFisher Scientific Inc . ) . For analysis of Ysp2 protein expression , 10 OD600 units of GFP-Ysp2–expressing cells were prepared and proteins extracted as above . To detect GFP-Ysp2 , immunoblots were incubated with 1:2 , 000 anti-GFP antibody ( Sigma-Aldrich Chemicals ) followed by 1:10 , 000 anti-rabbit-HRP secondary antibody ( Promega , Madison , WI ) . GAPDH was detected using 1:10 , 000 anti-GAPDH antibody ( ThermoFisher Scientific Inc . ) followed with 1:10 , 000 anti-mouse-HRP secondary antibody ( Promega ) . We derive a rough estimate of the chance of finding cER at the cell cortex in yeast cells that lack the ability to tether the ER to the PM as follows . Assuming that a yeast cell has a volume of 65 μm3 ( radius = 2 . 5 μm ) [102 , 103] , we estimate the volume of a cortical shell defined by the reported distance ( 30 nm ) at which the ER is retained at the PM by tethers as 1 . 9 μm3 . As about 45% of the PM is associated with ER in WT cells ( Fig 1E and references [5 , 32] ) , the volume of the cortical shell that is occupied by ER is 0 . 9 μm3 . If this amount of ER were to become untethered , then it could be found anywhere in the total volume of the cell . Approximately 65% of the total cell volume is available for this purpose , i . e . , 42 μm3 , as the rest is occupied by the nucleus and organelles [104] . Thus , the random chance of finding the dispersed complement of cER anywhere in the cell , including the cell cortex , is 0 . 9/42 = 0 . 02 , or about 2% . | Almost half of the inner surface area of the yeast plasma membrane ( PM ) is covered with closely associated cortical endoplasmic reticulum ( ER ) . In yeast and human cells , it has been proposed that ER-anchored tether proteins staple the ER to the PM , creating membrane contact sites at which lipid transport between the ER and PM and membrane lipid synthesis are coordinately regulated , but the potential mechanisms are unclear . Here , we test this idea by creating yeast cells that lack all ER-PM tethers . We find that whereas the bidirectional transport of sterols between the ER and PM is unaffected in these cells , sterols within the PM are disorganized due to disruptions in phospholipid biosynthesis that alter PM lipid composition . In particular , we show that phosphatidylinositol-4-phosphate , a phospholipid needed for intracellular signaling and membrane trafficking , accumulates within the PM . Some of these defects can be rescued by reinstating membrane contacts via expression of an artificial tether . However , correction is also achieved without the creation of contacts by supplementing the growth medium with a precursor of membrane phospholipids . Based on these results , we propose that ER-PM contacts do not play a major role as physical conduits for lipid exchange but rather serve as regulatory interfaces to integrate lipid synthesis pathways . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"b",
"vitamins",
"chemical",
"compounds",
"endoplasmic",
"reticulum",
"cell",
"processes",
"organic",
"compounds",
"cholines",
"membrane",
"proteins",
"fungi",
"cellular",
"structures",
"and",
"organelles",
"lipids",
"chemistry",
"vitamins",
"sterols",
"cell",
"membranes",
"yeast",
"biochemistry",
"eukaryota",
"cell",
"biology",
"secretory",
"pathway",
"organic",
"chemistry",
"phospholipids",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"organisms"
] | 2018 | Endoplasmic reticulum-plasma membrane contact sites integrate sterol and phospholipid regulation |
Numerous gene fusions have been uncovered across multiple cancer types . Although the ability to target several of these fusions has led to the development of some successful anti-cancer drugs , most of them are not druggable . Understanding the molecular pathways of a fusion is important in determining its function in oncogenesis and in developing therapeutic strategies for patients harboring the fusion . However , the molecular pathways have been elucidated for only a few fusions , in part because of the labor-intensive nature of the required functional assays . Therefore , we developed a domain-based network approach to infer the pathways of a fusion . Molecular interactions of a fusion are first predicted by using its protein domain composition , and its associated pathways are then inferred from these molecular interactions . We demonstrated the capabilities of this approach by primarily applying it to the well-studied BCR-ABL1 fusion . The approach was also applied to two undruggable fusions in sarcoma , EWS-FL1 and FUS-DDIT3 . We successfully identified known genes and pathways associated with these fusions and satisfactorily validated these predictions using several benchmark sets . The predictions of EWS-FL1 and FUS-DDIT3 also correlate with results of high-throughput drug screening . To our best knowledge , this is the first approach for inferring pathways of fusions .
Gene fusions resulting from chromosome rearrangements and chimeric RNAs generated by trans-splicing or read-through events play important roles in cancer . Recently , the advent of massively parallel sequencing has accelerated the rate of the discovery of fusions across multiple cancer types [1] . Fusions , especially druggable kinase fusions , may serve as therapeutic targets in cancer [2] . However , many identified fusions are not druggable , such as EWS-FLI1 [3] and C11orf95-RELA [4] . In addition , drug resistance to fusion-targeted therapies , such as resistance to imatinib in chronic myeloid leukemia ( CML ) , may develop in some patients owing to mutations that render a drug unable to bind to its targeted fusion or to the activation of compensatory pathways [5] . A better understanding of the molecular pathways of fusions would facilitate the development of therapeutic strategies for patients who harbor these fusions . However , owing to the labor-intensive nature of the validation assays required , few studies have investigated the pathways of novel fusions [6] . Therefore , computational methods for predicting the pathways of fusions would greatly help understand their functions in oncogenesis and develop therapeutic strategies . Numerous methods to annotate the functional impact of point mutations by mutation rate , reading frame , and evolutionarily conserved regions have been developed [7] . In particular , some methods incorporate pathway-level information and gene expression data to assess the functional impact of mutated genes [8] . These pathway-based methods suggest that important mutated genes deregulate the expression levels of their interacting partners and pathways , whereas passenger mutations have little impact . However , these approaches may not be applicable to fusions , whose interacting partners and pathways underlying their oncogenesis remain obscure . Several computational approaches that can prioritize fusion drivers in sequencing data have been proposed recently . Wu et al . [9] developed a fusion centrality metric to score fusions under the assumption that a fusion is more likely to be an oncogenic driver if its parental genes act like hubs in a molecular network . Two other approaches , Oncofuse [10] and Pegasus [11] , use supervised learning methods to identify fusion drivers by a feature space composed of protein domains , reading frame annotations , and protein-protein interaction partners . However , these approaches cannot elucidate the pathways of a fusion . Several computational approaches for predicting molecular interactions of proteins on the basis of domain information have been developed [12] . A fusion retains some of the functional domains of its parental genes [13] . Thus , we reason that molecular interactions of a fusion can be inferred given the composition of its protein domains . Latysheva et al . [14] recently unified all the protein interactions of the parental proteins of a fusion and found a gene fusion may rewire the protein interaction network in cancer through connecting proteins that did not previously interact in the network . But , they did not consider that a fusion may not inherit some protein interactions from its parental proteins because of losing some of their protein domains . In addition , a very recent study [15] presented a method , “ChiPPI” , to identify preserved protein-protein interactions of fusions using the domain-domain co-occurrence scores . Yet , this method does not help uncover the pathways of a specific fusion . Herein , we propose a domain-based network approach to infer the molecular interactions and pathways associated with a fusion and explore potential therapeutic targets in these pathways . We demonstrated the capabilities of our approach by applying it to the BCR-ABL1 fusion and two undruggable fusions , EWS-FL1 and FUS-DDIT3 .
Our proposed domain-based network approach is illustrated in Fig 1 . First , a fusion protein consists of the domains of its parental proteins , therefore we hypothesized that the fusion protein retains some of the molecular interactions of its parental proteins . Given the fusion’s domain composition , we can predict the protein-protein or protein-DNA interactions of the fusion ( Fig 1A ) . Protein-protein interactions result primarily from the binding of a modular domain of one protein to the domains of another protein [16] . The protein-protein interaction partners of the fusion’s parental proteins , whose domains interact with the domains of the fusion , are predicted to be the protein-protein interaction partners of the fusion . In addition , some fusions act as transcription factors to deregulate gene transcriptions through protein-DNA interactions . Because protein-DNA interactions mainly occur when a transcription factor composed of DNA-binding domains binds to a motif within promoters of its target genes to regulate their expression , we simply assumed that a fusion containing DNA-binding domains of its parental genes would be able to deregulate the transcriptional regulation interactions of its parental genes . A set of protein-protein interactions , a set of protein domain-domain interactions , and a set of transcriptional regulation interactions were used to predict molecular interactions of fusions ( S1 Text ) . Second , we reasoned that genes that are close to these predicted molecular interactions in the network would be functionally associated with the fusion , a principle known as “guilt by association” [17] ( Fig 1B ) . Our approach uses the random walk with restart [18] to calculates the functional relatedness of each gene with the fusion ( termed fusion-association score ) on the basis of the relative location of each gene to the predicted molecular interaction partners of the fusion in a compiled gene network ( S1 Text ) . On the basis that the hypothetical random walkers diffuse along the links of the network from the predicted molecular interaction partners , genes that are close to the predicted molecular interaction partners will be those most often visited by the random walkers during their random walks . Given a network with n nodes ( i . e . , genes in the gene network ) , the random walk with restart is defined as: pt+1= ( 1‑γ ) Tpt+γp0 ( 1 ) where p0 is the initial probability vector in which equal probabilities are assigned to the starting nodes ( i . e . , the predicted molecular interaction partners ) , pt is the probability vector containing the probabilities of the nodes at step t , γ is the restarting probability , and T is the transition matrix , which is a column-normalized adjacency matrix of the network . Starting from the set of nodes in the network , the walker will iteratively move from the current nodes to randomly selected neighbor nodes or return to the starting nodes . When iteratively reaching stability ( i . e . , when the change between pt and pt+1 is below 10e-30 ) , the probability vector can present the association scores of all genes to the starting genes . Thus , genes with higher association scores are also more associated with the fusion . Third , we considered pathways that enrich genes with high fusion-association scores are also associated with the fusion ( Fig 1C ) . We ranked all the genes on the basis of their fusion-association scores . Gene set enrichment analysis ( GSEA ) [19–20] , which can evaluate the genes of a pathway for their distribution in the ordered gene list , was then used to identify pathways functionally associated with a fusion ( termed GSEA association analysis ) . However , because the gene network used in the prediction contains molecular interactions across different cellular statuses , not all predicted associated pathways are deregulated by a fusion in a specific cellular condition . By leveraging gene expression data from samples that have or lack a fusion , we can identify pathways that are deregulated by the fusion ( Fig 1D ) . The deregulation level of an associated pathway can be also determined using GSEA ( termed GSEA deregulation analysis ) . The truncated product method [21] was used to combine the p values of each pathway generated from the GSEA association and GSEA deregulation analyses to identify pathways that are both highly associated with the fusion and significantly deregulated by it in a specific cellular condition . In the truncated product method , the product score W of the two p values ( pi ) that do not exceed a fixed τ value ( τ was set to 0 . 01 for both p values ) can be calculated as: W=∏i=12piI ( pi≤τ ) ( 2 ) where I ( . ) is the indicator function . The probability of W for w<1 can be evaluated by conditioning on the number , k , of the pi’s less than τ: Pr ( W≤w ) =∑k=12Pr ( 2k ) ( 1‑τ ) 2‑k ( w∑s=0k‑1 ( klnτ‑lnw ) ss ! I ( w≤τk ) +τkI ( w>τk ) ) ( 3 ) Literature co-citation data have been widely used to infer gene-gene and gene-disease functional associations [22–23] . Therefore , we compiled two types of benchmark gene sets from literature citation data to comprehensively evaluate each fusion prediction . First , we collected 378 genes cited in at least two papers related to BCR-ABL1 , 29 genes cited in at least two papers related to EWS-FLI1 , and 11 genes cited in at least one papers related to FUS-DDIT3 ( As only few literature citations of FUS-DDIT3 were found , a lower criteria was applied here ) to evaluate the predictions of genes functionally associated with these fusions . We downloaded all PubMed identification numbers of articles related to each fusion . These numbers were then cross-referenced with the gene citation information from Entrez Gene ( ftp://ftp . ncbi . nih . gov/gene/ ) , which is composed of genes and corresponding cited literature . Second , using the similar method , we collected 416 genes cited in at least two papers related to CML , 98 genes cited in at least two papers related to Ewing’s sarcoma , and 61 genes cited in at least one paper related to myxoid liposarcoma to evaluate whether fusion-associated genes in our prediction play a role in the oncogenesis of the specific cancer type . We also used a set of 328 cancer-related genes compiled from KEGG cancer pathways to evaluate the association between cancer pathways and the three fusions . To evaluate our predictions of therapeutic targets in fusion-associated pathways , we collected 68 target genes of drugs that have been tested in clinical trials or used for the treatment of CML and 67 drug target genes for Ewing sarcoma ( S1 Table and S3 Table ) . These target genes were manually collected from the literature and available public databases ( detailed in S2 Text and S3 Text ) . Furthermore , we also evaluated our predictions using target genes of sensitive compounds identified in high-throughput screening for two Ewing’s sarcoma cell lines ( TC32 and TC71 ) and three myxoid liposarcoma cell lines ( MLS-1765-92 , MLS-402-91 , and MLS-DL221 ) . The material and method of high-throughput screening are detailed in the S1 Text . We identified 76 sensitive compounds for Ewing cell lines and 48 sensitive compounds for myxoid liposarcoma cell lines . With the incorporation of known drug target data , we respectively found 60 of the 76 sensitive compounds for Ewing cells and 38 of the 48 sensitive drugs for the myxoid liposarcoma cells have known target genes . Totally , we have 197 targets of sensitive compounds for Ewing cells and 161 targets for myxoid liposarcoma cells ( S4 Table and S5 Table ) .
A variety of breakpoints in BCR and ABL1 rearrangements generate BCR-ABL1 chimeric proteins with different domain compositions , of which the three major variants ( p185 , p210 , and p230 ) occur in different types of leukemia [24] . These variants may be associated with tissue-specific spatial organization of genomes [25] . The ABL1 portion of all three variants contains tandem SRC homology , the tyrosine kinase domains , SH3 binding sites , a DNA-binding domain , and an actin-binding domain [24] . However , the BCR portions of the three variants differ . p230 contains the largest amount of BCR sequence , including a calcium-binding domain , a Gap-Rac domain , a coiled-coil oligomerization domain , a serine/threonine kinase domain , a Dbl homology domain , and a pleckstrin homology domain . p210 and p185 both lose the Gap-Rac domains , but p185 also loses the Dbl homology and pleckstrin homology domains . We first applied our approach to the p210 BCR-ABL1 , which is the most commonly associated with CML . Our approach first predicted the protein-protein interaction partners of p210 BCR-ABL1 . Many proteins that are known to physically interact with BCR-ABL1 , such as GRB2 , CRKL [26] , and SOCS1/3 [27] , were in our prediction ( S2 Text ) . This supports our hypothesis that a fusion retains some molecular interactions of its parental genes . Next , we used these predicted protein interaction partners to infer BCR-ABL1-associated pathways . GSEA association analysis revealed that several known BCR-ABL1 pathways are highly associated with BCR-ABL1 in our prediction ( Fig 2A ) , such as imatinib pathway ( Fig 2B shows its GSEA association plot ) , JAK-STAT signaling pathway [26] , DNA damage pathway [28] , Hedgehog pathway [29] , and p53 pathway [30] . We also analyzed the gene expression data of three CML cell lines and their imatinib-treated cells [31] and examined the deregulation of these pathways using GSEA . The combination analysis of the GSEA association and GSEA deregulation analyses shows that these pathways are both highly associated with BCR-ABL1 and significantly deregulated upon its inhibition ( Fig 2A ) . Our approach also predicted most of 26 Wnt/Ca+/NFAT pathway genes identified by RNAi-based screen with imatinib in CML cells [32] are associated with BCR-ABL1 ( Fig 2C ) . Specifically , CAMK2B , one of the top 5% of genes associated with BCR-ABL1 in our prediction , is a target of cyclosporin A , which can sensitize BCR-ABL-positive leukemia to BCR-ABL inhibitors [32] . This implies that our approach can help identify therapeutic targets associated with a fusion ( detailed in the following section ) . In addition , we also collected several literature-based benchmark gene sets to comprehensively evaluate our prediction , including BCR-ABL1 related genes , CML related genes , and cancer pathway genes ( Fig 2C ) . These evaluations using the ROC analysis ( Fig 2C ) and the other two statistical tests ( S2 Text ) all showed that most of these benchmark genes were highly associated with BCR-ABL1 in our prediction . Furthermore , our predictions of BCR-ABL1 were also correlated with several data-driven gene signatures associated with BCR-ABL1 and CML ( S1 Fig ) . These results all indicate that our approach can successfully predict cancer-related genes and pathways that are associated with BCR-ABL1 . Furthermore , we also compared our predictions of the three BCR-ABL1 variants . We found that the three variants had similar numbers of predicted protein interactions ( Fig 3A ) and the three variants had almost the similar pathway predictions ( Fig 3B ) . Prediction evaluation using the 328 cancer pathway genes ( Fig 3C ) also indicated that these variants would have almost the same oncogenic impact . Similar results were observed in evaluations using different benchmark sets ( S2 Fig ) . Some studies also have shown that these three variants are equally potent in inducing a CML-like disease in transplanted mice [33] . Inhibition of BCR-ABL1 fusion by imatinib has proven to be a very successful treatment for CML with the BCR-ABL fusion . However , some patients develop imatinib resistance owing to the emergence of BCR-ABL1 point mutations . Several second-generation drugs that target both BCR-ABL1 and its associated pathways thus have been developed for the treatment of imatinib-resistant CML [5] . Therefore , we also used p210 BCR-ABL1 as an example to illustrate the capability of our approach to identify therapeutic targets associated with a fusion . In our literature review , we collected 68 target genes of 24 drugs that have been in clinical trials or used for the treatment of CML ( S1 Table ) . We successfully predicted that most of these 68 genes are highly associated with BCR-ABL1 ( Fig 2C ) . This indicates that our approach could help develop therapeutic strategies that target the pathways associated with a fusion . Therefore , we also mapped the drug target data from the DGIdb database [34] to the top 10% of genes associated with BCR-ABL1 in our prediction . We identified some of these compounds , such as ruxolitinib , regorafenib , and arsenic trioxide , have been shown to have an anti-cancer effect in CML cells or other cancer cells and could be potential treatments for CML patients ( Table 1 ) . Ruxolitinib , a JAK1/2 inhibitor , is used for the treatment of intermediate or high-risk myelofibrosis . Studies in vitro and in vivo support the use of ruxolitinib for the treatment of CML [35–36] . Regorafenib , a multikinase inhibitor , has been approved for the treatment of advanced gastrointestinal stromal tumors and metastatic colorectal cancer [37] . Regorafenib can target several BCR-ABL1–associated genes , such as KIT . Some studies showed that KIT signaling governs differential sensitivity to tyrosine kinase inhibitors in mature and primitive CML progenitors [38] . Arsenic trioxide has been approved for the treatment of acute promyelocytic leukemia [39] , and recent work suggests that arsenic trioxide is also a treatment option for CML [40] . We also integrated the gene expression data of three CML cell lines and their imatinib-treated CML cells [31] with the top genes associated with BCR-ABL1 in our prediction to prioritize targets whose inhibition can have synergistic effects with imatinib . About 500 genes were significantly up-regulated in imatinib-treated CML cells ( fold change >1 . 5 and p < 0 . 05 ) , and 50 of them were in the top 10% of genes in our prediction ( S2 Table ) . Some of the 50 genes are potential therapeutic targets . For example , inhibition of JAK2 can overcome imatinib drug resistance in CML [35] . Inhibition of CSNK2A2 by CX-4945 also exhibits anti-tumor activity in chronic lymphocytic leukemia [48] . BCL6 , which has been shown to be up-regulated in response to treatment with imatinib , represents a novel defense mechanism enabling leukemia cells to survive despite imatinib treatment [31] . In addition , studies also showed that CBL-B is required for the leukemogenesis mediated by BCR-ABL through the negative regulation of bone marrow homing [49] and that CML with the CBL-B mutation is resistant to imatinib [50] . Although some of these predicted targets in our analysis are currently undruggable , developing RNAi-based therapies would enable us to target these genes [51] . Sarcomas have greater than 50 histological subtypes , which arise in bone , cartilage or connective tissues . Currently , standard chemotherapy , radiation , and surgery are the only available treatments for most of sarcomas . Thus , alternative therapeutic strategies are urgently needed to improve survival rate and quality of life for sarcoma patients . Several types of sarcomas are driven by the presence of specific fusion mutations , such as SYT-SSX in synovial sarcoma , EWS-FLI1 in Ewing’s sarcoma , FUS-DDIT3 in myxoid liposarcoma [3] , and EWS-WT1 in Desmoplastic Small Round Cell Tumor ( DSRCT ) [52] . These fusions can produce tumor-specific proteins; thus are potential targets for the development of specific therapies for these sarcomas . However , most of these fusions are transcriptional regulators , making them more challenging as drug targets [2] . Understanding the pathways that are associated with these fusions would reveal alternative therapeutic strategies . In this work , we also applied our approach to two undruggable sarcoma fusions , EWS-FLI1 and FUS-DDIT3 . We successfully predicted the pathways that are known to be functionally associated with these two fusions , for instance , WNT [53] , IGF1 signaling [54] , and PDGFR pathways [55] for EWS-FLI1 ( S3 Fig ) , and adipocytokine signaling [56] , DNA damage [57] , NF-kB pathways [58] , and FGFR pathway [59] for FUS-DDIT3 ( S5 Fig ) . Several pathways that were recently found to be associated with EWS-FLI1 are also identified by our approach , such as chromatin remodeling [60] , splicing pathways [61] , and CRM1-dependent nuclear export pathway [62] ( S3 Fig ) . Our predictions were also satisfactorily validated by several literature-based benchmark sets ( Fig 4 , S3 Text , and S4 Text ) . In addition , several data-driven gene signatures associated with Ewing’s sarcoma and EWS-FLI1 were also correlated well with our predictions ( S4 Fig ) . These results indicate that our approach can successfully identify most of known genes and pathways associated with the two fusions . High-throughput screening has been widely used in drug-discovery processes to rapidly identify compounds that are active againsts of particular pathways in tumours [63] . Therefore , we also compared our predictions of two fusions respectively with the high-throughput screening results in two Ewing’s sarcoma and three myxoid liposarcoma cell lines . The known target genes of the identified sensitive compounds in the screening were used to evaluate our predictions ( S4 Table and S5 Table ) . The ROC analysis of the evaluation shows that our predictions correlate well with the screening results ( Fig 4 ) . We also found that 51 of the 60 ( 85% ) sensitive compounds for Ewing cells have at least one target gene ranked among the top 10% of those predicted to be associated with EWS-FLI1 , and 37 of the 38 ( 97 . 37% ) sensitive compounds for the myxoid liposarcoma cells have at least one target gene ranked among the top 10% genes associated with FUS-DDIT3 . Specifically , some studies showed that Ewing sarcoma cells are sensitive to PARP1 inhibition [64–65] , and our high-throughput screening results also showed that the two Ewing cells are sensitive to a PARP1 inhibitor , BMN 673 . Our prediction also successfully identify PARP1 is one of top 5% genes that are functionally associated with EWS-FLI1 in the prediction . In addition , we also collected target genes of drugs that have been tested in clinical trials or used for the treatment of Ewing sarcoma to evaluate our prediction of EWS-FLI1 . The evaluation shows that most of these target genes are functionally associated with EWS-FLI1 in our prediction ( Fig 4A ) . Furthermore , Myxiod liposarcoma has been shown to be high sensitive to a novel chemotherapeutic agent , Trabectedin [66] . The recent data suggested Trabectedin may block the transactivation of FUS-DDIT3 [67] . We also found that our prediction of FUS-DDIT3 was correlated with several data-driven signatures associated with Trabectedin ( S6 Fig ) . These results show that our approach can be used to identify therapeutic targets in the associated pathways of these two undruggable fusions and can be a way of in-silico drug screening . Other results of these two fusions are detailed in the S3 Text and S4 Text . The R package “FusionPathway” is available in the GitHub repository ( https://github . com/perwu/FusionPathway/ ) and runs in R environment . Installation of the R package is described in the GitHub page of “FusionPathway” . The scripts and data for generating the results of three examples demonstrated in this manuscript are also provided in the package . The general workflow of the package is shown in Fig 5 , in which we took BCR-ABL1 as an example . First , the basic input data includes two data frame objects: “GeneData” and “DomainData” . “GeneData” should contain Entrez gene IDs and gene symbols of the two parental genes whereas “DomainData” should contain Pfam protein domain IDs of the two parental genes , which are retained and lost in a fusion . Second , the FusionPathway function in the package was used to generate the association prediction of a fusion . Third , the outputs include four files: 1 . a list of predicted the molecular interaction partners of a fusion; 2 . an ordered gene list ranked by the fusion-association prediction; 3 . results of GSEA pathway association analysis; 4 . results of mapped drugs . Those top genes associated with a fusion can be selected using the values of rank percentage ( e . g . 10% ) , the top pathways associated with a fusion can be chosen using the p values of the GSEA association analysis , and potential drugs can be prioritized using the selected top genes . Fourth , by integrating gene expression data from samples that have or lack a fusion with our association prediction , we can identify pathways that are both highly associated with a fusion and significantly deregulated by it in a specific cellular condition . The fGSEA R package [20] ( also included in the package ) can be first used to do GSEA pathway deregulation analyses based on the gene list ordered by p-values of the differential gene expression analysis . The Combine_Pathway_Pvalues function can then be used to integrate pathway results of the GSEA association and GSEA deregulation analyses . We will make continual improvements to the package for its performance and usability .
In our domain-based network approach , we hypothesized that a fusion retains some of the functional domains and its cognate molecular interactions from its parental genes . Indeed , we found that most known BCR-ABL1 protein-protein interactions were inherited from BCR and ABL1 . Thus , the molecular functions of a fusion would be mainly associated with the functions of its parental genes . This implies that a fusion plays an important role in oncogenesis if at least one of its parental genes is associated with cancer . Some also believe that , rather than demolishing the entire machinery and creating a new version to trigger oncogenesis , cancer cells normally make modifications to existing cellular mechanisms [68] . Several recent studies support this view . Two studies [9 , 14] found that the parental genes of many known cancer fusions are also cancer-related and act as hubs in molecular networks . Other studies also revealed that important fusions in cancer normally retain functional domains associated with oncogenesis from their parental genes [13] . However , a fusion may have novel functions and molecular interactions that are absent in its parental genes because of the shuffling of different active sites and binding domains or the generation of new domain combinations [69] . Several computational approaches can predict the novel molecular interactions of proteins on the basis of multiple domain combinations [70] . These approaches can be integrated into our model to predict the novel protein interactions of a fusion based on its novel domain combinations . Our results reveal that our approach can help identify potential therapeutic targets in the pathways associated with a fusion . However , not all of the top genes associated with a fusion in our prediction can be therapeutic targets because their associations with the fusion may relate to different types of phenotypic effects [71] . Therefore , integrating other high-throughput data , such as gene expression data and RNAi screening , with the predicted top genes can help prioritize potential therapeutic targets . For instance , by integrating gene expression data , we successfully identified genes whose inhibition can sensitize imatinib-resistant CML cells to treatment . In addition , RNAi screening has been used to identify critical genes that control cancer-related phenotypes without using any prior biological information [72] . However , when no prior biological information is used , RNAi screening must be used to assess all genes . In contrast , integration of top genes in our prediction with RNAi screening can greatly reduce the search space . With the incorporation of drug target data , our approach also can help identify compounds that may have action in pathways associated with a fusion . However , some of these identified compounds may be ineffective in the treatment of patients harboring the fusion [73] . First , the drug-binding affinities are target-dependent . In addition , the mechanisms of action of some drugs are unclear . Finally , most drugs act against multiple targets , and inhibition of these targets may lead to different types of phenotypic effects , both positive and negative [71] . Therefore , evaluating the possible therapeutic effects of the identified drugs is difficult . However , our approach could serve as a quick way to initially screen a small number of potential drugs subjected to further evaluation , such as high-throughput screening . The good correlations between our predictions of two sarcoma fusions and the high-throughput screening results support this view . This indicates that integration of our approach predictions can help reduce the search space in high-throughput screenings [74] . Moreover , if the targets of some drugs that have been already approved by the U . S . Food and Drug Administration are highly ranked in the prediction , such drugs could be repositioned for the treatment of cancers harboring specific fusions . Diverse drug resistance mechanisms , including pathway-dependent mechanisms ( e . g . , target reactivation through secondary mutation , downstream activation , or bypass activation ) and pathway-independent mechanisms ( e . g . , tumor microenvironment perturbation ) , remain major problems for targeted therapies [75] . Combinations of drugs targeting the associated pathways may prevent drug resistance [76] . Our results of BCR-ABL1 suggest that our approach may help elucidate pathway-dependent mechanisms of resistances to those kinase fusion-targeting therapies and develop strategies to overcome the resistances . Very recently , a method , ChiPPI , was proposed to identify preserved protein-protein interactions of fusion proteins [15] . We found there are two major differences between our study and ChiPPI . First , ChiPPI only predicts protein-protein interactions of fusions while our method can predict both of protein-protein interactions and protein-DNA interactions of fusions . Thus , our method could more accurately uncover molecular mechanisms of those fusions that act as transcription factors to deregulate pathways mainly through protein-DNA interactions . Second , rather than doing pathway analysis of each fusion , ChiPPI performed a pathway enrichment analysis of all the predicted interactors of all the fusions in the three major disease types , leukemia/lymphoma , sarcoma or solid tumors , to identify the over-presented pathways in each of the three cancer types . In contrast , our method used the guilt-by-association method to identify pathways that are associated ( either directly and undirectly interact ) with an individual fusion . Our method can help understand molecular mechanisms of a specific fusion and explore therapeutic targets in these pathways for patients harbouring the fusion . Both our method and CHiPPI predicted which protein interactions of parental proteins will be kept in a fusion given the domain compositions of a fusion . But , our method used a set of curated protein domain-domain interactions ( S1 Text ) to predict protein interactions of fusions while ChiPPI predicts interactions based on domain-domain co-occurrence scores that were calculated de novo based on all collected interactions in the BioGrid database [77] . In addition , our method predicted protein interactions of fusions from a set of protein-protein interactions compiled from several expert-curated databases ( S1 Text ) while ChiPPI made predictions based on protein interactions from BioGrid . We here used 33 known protein interactors of BCR-ABL1 ( S2 Text ) that were compiled from literature to compare BCR-ABL1 ( p210 ) predictions of our method and ChiPPI . We found that our method predicted 30 of the 33 known protein interactions while ChiPPI predicted 24 the known interactions . However , we need to emphasize that it is difficult to evaluate prediction performance between our method and ChiPPI because it requires robust positive and negative interactions of a fusion , which does not currently exist . Our approach is not without limitations . Because it is based on incomplete molecular interaction data , our approach cannot be used to determine the biological functions of fusions whose parental genes or functional domains have not yet been well characterized . In addition , our approach is not able to predict novel molecular interactions that are absent in its parental genes because of the shuffling of different active sites and binding domains or the generation of new domain combinations or other mechanisms . However , our approach will improve over time as more data are generated . Regardless , our results indicate that our approach can be an effective method for inferring the pathways of a fusion and identifying potential therapeutic targets in these pathways . This will greatly help develop therapeutic strategies for patients who harbor undruggable fusions and for those whose disease is resistant to fusion-targeted therapies . | We present a computational framework , FusionPathway , to infer the oncogenesis pathways of a fusion and help develop therapeutic strategies in these pathways for patients harboring the fusion . In this work , we successfully validated the capabilities of this approach through its application to the well-studied BCR-ABL1 fusion and two undruggable fusions in sarcoma , EWS-FL1 and FUS-DDIT3 . Especially , the predictions of EWS-FL1 and FUS-DDIT3 correlate well with results of high-throughput drug screening in sarcoma cells . Therefore , FusionPathway can be an effective method to infer pathways and potential therapeutic targets that are associated with those undruggable fusions . Our results of BCR-ABL1 also suggest that FusionPathway may be able to help elucidate pathway-dependent mechanisms of resistances to those kinase fusion-targeting therapies and develop strategies to overcome the resistances . In addition , the developed R package of FusionPathways ( https://github . com/perwu/FusionPathway/ ) can help people easily apply our approach to study other important fusions in cancer . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"protein",
"interactions",
"cancer",
"treatment",
"protein",
"interaction",
"networks",
"cancers",
"and",
"neoplasms",
"oncology",
"ewing",
"sarcoma",
"network",
"analysis",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"protein-protein",
"interactions",
"proteins",
"high",
"throughput",
"screening",
"proteomics",
"sarcomas",
"molecular",
"biology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"biochemistry",
"cell",
"biology",
"protein",
"domains",
"biology",
"and",
"life",
"sciences",
"cell",
"fusion"
] | 2018 | FusionPathway: Prediction of pathways and therapeutic targets associated with gene fusions in cancer |
Circadian clocks are aligned to the environment via synchronizing signals , or Zeitgebers , such as daily light and temperature cycles , food availability , and social behavior . In this study , we found that genome-wide expression profiles from temperature-entrained flies show a dramatic difference in the presence or absence of a thermocycle . Whereas transcript levels appear to be modified broadly by changes in temperature , there is a specific set of temperature-entrained circadian mRNA profiles that continue to oscillate in constant conditions . There are marked differences in the biological functions represented by temperature-driven or circadian regulation . The set of temperature-entrained circadian transcripts overlaps significantly with a previously defined set of transcripts oscillating in response to a photocycle . In follow-up studies , all thermocycle-entrained circadian transcript rhythms also responded to light/dark entrainment , whereas some photocycle-entrained rhythms did not respond to temperature entrainment . Transcripts encoding the clock components Period , Timeless , Clock , Vrille , PAR-domain protein 1 , and Cryptochrome were all confirmed to be rhythmic after entrainment to a daily thermocycle , although the presence of a thermocycle resulted in an unexpected phase difference between period and timeless expression rhythms at the transcript but not the protein level . Generally , transcripts that exhibit circadian rhythms both in response to thermocycles and photocycles maintained the same mutual phase relationships after entrainment by temperature or light . Comparison of the collective temperature- and light-entrained circadian phases of these transcripts indicates that natural environmental light and temperature cycles cooperatively entrain the circadian clock . This interpretation is further supported by comparative analysis of the circadian phases observed for temperature-entrained and light-entrained circadian locomotor behavior . Taken together , these findings suggest that information from both light and temperature is integrated by the transcriptional clock mechanism in the adult fly head .
Organisms on Earth have evolved an internal timekeeping system , or circadian clock ( circa = about , diem = day ) , that allows them to both respond to and predict changes in the 24-h environmental day . Much has been learned about the genes involved in this precise , 24-h molecular timekeeping mechanism in the fruit fly Drosophila melanogaster ( for a recent review see [1] ) . The fly clock is composed of intracellular feedback loops: The proteins Clock ( CLK ) and Cycle ( CYC ) activate transcription of period ( per ) , timeless ( tim ) , vrille ( vri ) , and PAR-domain protein 1 ( Pdp1 ) . Subsequently , proteins encoded by the latter four genes either suppress or activate CLK and CYC [2–8] . Feedback in these regulatory loops is thought to oscillate due to timed changes in the stabilities and subcellular localizations of component proteins , especially Period ( PER ) and Timeless ( TIM ) [9 , 10] . The fly molecular clock is aligned to the environment through Zeitgebers ( “time givers” ) , the most notable being the daily light/dark cycle . This is mediated by the light-dependent degradation of the TIM protein [11 , 12] . Cryptochrome ( CRY ) , a blue light photoreceptor in the family of flavoproteins , has been shown to associate with TIM during the light phase of the circadian day , resulting in ubiquitination and degradation of TIM by the proteasome and ultimately relieving inhibition of CLK-mediated transcription [13–15] . In addition , a second pathway of light entrainment in the pacemaker neurons is defined by signals from visual organs that may impact TIM in a CRY-independent manner [13 , 16] . Light is the best understood Zeitgeber , but other factors , such as daily changes in temperature [17–20] and social behavior [21] , can act as inputs to the fly circadian clock . Although the fly clock is temperature-compensated over a wide range of constant physiological temperatures , it has been known for several decades that eclosion in Drosophila pseudoobscura can entrain cycling temperature changes [20] . Further , it was shown in this species that temperature step-ups , step-downs , and pulses result in accompanying phase shifts in behavior [22] . Locomotor activity behavior in D . melanogaster can be entrained to temperature cycles of as little as 3 °C [18] . The locomotor activity rhythms of arrhythmic clock mutants ( per0 , tim01 , ClkJrk , cyc0 ) can be driven by temperature [19] . However , these “clock-less” mutants do not truly entrain as there is no anticipation of the temperature transitions , and rhythmicity does not persist when they are released into constant conditions . Interestingly , the locomotor activity of wild-type flies can be entrained to temperature cycles during constant light , a condition that would normally result in arrhythmicity [17 , 19 , 23] . There is anticipation of temperature transitions , but as with the arrhythmic clock mutants , locomotor activity behavior becomes arrhythmic when the temperature cycle is removed . Molecularly it has been shown that short , high temperature heat pulses result in rapid downregulation of both PER and TIM proteins [24] . This results in a phase delay if the heat pulse is administered in the early night . However , a heat pulse given in the late night does not result in a phase advance , as is the case with a light pulse given at this time . This is thought to be due to a rapid increase in PER and TIM production after the initial downregulation , ultimately resulting in a constant period [24] . It is not clear if the molecular responses triggered by abrupt heat pulses also play an important role in the entrainment of the molecular clock circuits to environmental temperature cycles . PER and TIM proteins oscillate during temperature entrainment in constant darkness ( dark/dark or DD ) , and these oscillations are maintained during constant conditions following entrainment [13] . It thus follows that temperature acts on at least some of the same molecular components of the circadian clock as light . Temperature cycles can also drive PER and TIM oscillations during constant light , a condition that , as mentioned earlier , normally results in behavioral and molecular arrhythmicity [17 , 23] . In this report , we examine temperature as a Zeitgeber for the circadian clock and ask whether information from temperature is relayed through the same molecular circuits as light . In nature , the maxima and minima of solar irradiation and environmental temperature are offset ( Figure 1 ) [25] . Sunrise generally coincides with the coolest part of the day and maximum solar irradiance at noon precedes the temperature maximum in the late afternoon . The divergent phases and waveforms of the environmental light and temperature profiles could in principle be represented by separate Zeitgeber-specific oscillators , or they could be integrated by a single oscillator capable of synchronizing to both photocycles and thermocycles . By generating genome-wide transcriptional profiles during temperature cycles and subsequent constant conditions , we show that there are two distinct responses to temperature: a clock-independent , temperature-driven response and a clock-dependent , circadian response . Temperature-entrained circadian transcript profiles show a much higher degree of overlap with light-entrained circadian transcript profiles than do temperature-driven responses . Further , the mutual phase relationships among transcripts oscillating in response to both photo- and thermocycles are maintained in both conditions . Thus , many features of the circadian expression program emerge independently from the precise nature of the environmental Zeitgeber . Moreover , the molecular phases associated with separate photocycle and thermocycle entrainment suggest synergistic synchronization by the environmental light and temperature profiles found under most natural conditions .
Genome-wide transcript profiles for the heads of temperature-entrained flies were determined in four 12-point time course experiments conducted in constant darkness; three time courses ( two for wild-type and one for arrhythmic mutant tim01 flies ) consisting of a day-long 12-h 18 °C/12-h 25 °C ( cold/ambient or CA ) thermocycle plus a subsequent day of constant 25 °C ( ambient/ambient or AA ) and one time course ( for wild-type flies ) spanning the first two days of constant conditions ( see Materials and Methods ) . In order to identify high-confidence , 24-h periodic gene expression , we compared the distribution of oscillatory statistics obtained from Affymetrix high-density oligonucleotide arrays to a permutation null model ( see Materials and Methods; [26 , 27] ) . For each probe set on the arrays , we determined the 24-h spectral power and the probability of observing an equivalent or higher score from a genome-wide set of randomly permuted profiles . Analysis data are made available at http://biorhythm . rockefeller . edu . We then determined the number of selected 24-h periodic genes as a function of the threshold p-value or the associated false discovery rate ( FDR ) as illustrated by the graphs in Figure 2 . These analyses were performed for various combinations of the new datasets representing temperature entrainment and of previously described datasets [27–29] representing light entrainment . Our analysis method emphasizes coherence of phase and period length but does not penalize inter-experimental variation in amplitude [26] . We have used this method to demonstrate that analysis of combinations of independently obtained time course microarray datasets representing the same or a similar environmental protocol allows for the detection of periodic expression programs with improved resolution [26 , 27] . The quantitative differences between the observed 24-h periodic expression programs are best visualized in Figure 2A , where an arithmetic scale is used . It is clear that a 2-d wild-type dataset representing a CA environmental temperature cycle shows a much broader impact on global 24-h periodicity than an equivalent dataset representing a 12-h light/12-h dark ( LD ) cycle ( e . g . , 326 versus 42 periodic transcript profiles at FDR 0 . 2; see Figure 2A and 2B ) . This result could be explained by either a temperature-driven , clock-independent effect or by a thermocycle-specific , clock-dependent effect . Additional comparative analyses help to distinguish between these two possibilities . First , consider expanded 3-d versions of the CA and LD datasets that each also include a 1-d time course in the same format obtained from arrhythmic tim01 flies . Given that both behavioral and molecular circadian rhythms appear to be abrogated by the tim01 mutation [27 , 30] , the additional time course data may represent temperature- or light-driven , but not clock-dependent , expression profiles . The enhanced difference in 24-h periodicity between thermocycle and photocycle conditions that results from inclusion of the tim01 data ( e . g . , 939 versus 72 periodic transcript profiles at FDR 0 . 2; Figure 2A and 2C ) indicates that most of the thermocycle-associated transcript rhythms are simply temperature-driven independently from the clock . Second , compare the properties of 24-h periodic transcript profiles of 4-d datasets representing the same constant conditions ( 25 °C and constant darkness ) after either temperature entrainment ( AA ) or light entrainment ( DD ) . The circadian programs detected after entrainment to temperature and light have very similar properties ( Figure 2A and 2D; see also the section Defining a Set of Clock-Dependent Transcripts , below ) , supporting the hypothesis that the increased 24-h periodicity found in the context of an environmental temperature cycle is due to clock-independent temperature-driven regulation rather than to circadian rhythms that specifically require temperature entrainment . The broad temperature-driven response observed in the presence of a temperature cycle could in principle interfere with the circadian expression program . This issue is addressed by analysis of a 6-d dataset combining 2 d of temperature entrainment and 4 d of subsequent constant conditions ( CA/AA ) and an equivalent 6-d dataset representing light entrainment and subsequent constant conditions ( LD/DD ) . Both of these programs indicate more high-quality daily transcript oscillations than are found separately for the 2-d ( CA or LD ) or 4-d ( AA or DD ) subsets ( Figure 2A and 2C–2E ) . Even the sum total of high-quality daily transcript oscillations from the 2-d and 4-d subsets is considerably less than the number observed for the integrated 6-d sets ( e . g . , 33 for 2x CA plus 27 for 4x AA versus 212 for 2x CA/4x AA and 13 for 2x LD plus 20 for 4x DD versus 75 for 2x LD/4x DD at FDR 0 . 05 , Figure 2A and 2C–2E ) . This suggests that , in general , circadian expression profiles are not dramatically altered by temperature-driven ( or light-driven ) effects ( Figure 2A , 2B , 2D , and 2E ) . Comparison of the 24-h periodicity found in the 6-d wild-type CA/AA and LD/DD sets versus the 4-d wild-type AA and DD sets and the 3-d wild-type plus tim01 CA and LD sets further illustrates the magnitude of the temperature-driven response ( Figure 2A , 2C , and 2E ) . There is no reason to assume that purely circadian expression profiles are more prevalent or prominent in CA versus LD conditions , yet inclusion of 2 d of CA data with the AA set has a bigger impact on 24-h periodicity than the inclusion of 2 d of LD data with the DD set , suggesting that temperature-driven regulation may be responsible . In addition , although more extensive 24-h periodic expression is generally detected for larger datasets [26 , 27] , the daily expression program found for the 3-d wild-type plus tim01 CA dataset is comparable in size to that for the 6-d wild-type CA/AA dataset . In contrast , the 6-d wild-type LD/DD dataset shows much more extensive 24-h periodicity than the 3-d wild-type plus tim01 LD dataset , indicating that the presence of a daily thermocycle is the primary determinant of the number of observed 24-h transcript oscillations . Taken together , these analyses suggest that the circadian expression programs entrained by light entrainment and temperature entrainment have similar properties , whereas an environmental thermocycle directly evokes a global expression response that is considerably broader than that found for light-dependent or clock-dependent regulation . A core set of the most robustly temperature-driven transcripts was identified based on wild-type and tim01 thermocycle expression profile data . In order to be included , transcripts had to meet several noise filters and show a highly significant 24-h Fourier component as well as a significant 24-h autocorrelation ( see Materials and Methods ) . The phasegram for the resulting set of 164 temperature-driven transcripts in both wild-type and tim01 flies is shown in Figure 3A . The majority of these transcripts respond with a simple pattern of either activation during the warm phase and repression during the cold phase or vice versa . Further , most of this response is lost in constant conditions following temperature entrainment ( Figure 3A ) or during and after light entrainment ( unpublished data ) , emphasizing that the response is driven and not circadian . Figure 3B shows the average peak phases of the temperature-driven transcripts across data from wild-type and tim01 flies during temperature entrainment . Two trends are obvious from this analysis: ( 1 ) temperature-driven oscillations tend to peak around the middle of either the cryophase or the thermophase , and ( 2 ) more transcripts peak during the cryophase than during the thermophase . If the term CA0 is assigned to the onset of the cold temperature and CA12 to the onset of the ambient temperature , the majority of the transcripts have a phase of either CA5–8 ( toward the middle of the cryophase ) or CA18–20 ( toward the middle of the thermophase ) . Given our use of sine wave fits to estimate peak phases , the observed phase distribution is consistent with a majority of the expression profiles being directly positively or negatively temperature-driven with relatively little delay . Approximately three-quarters of the temperature-driven profiles peak in the cryophase , but the functional relevance of this preference is not directly obvious . The temperature-driven transcripts are representative of diverse biological functions ( Table S1 ) , including carbohydrate , amino acid , lipid/fatty acid , one-carbon compound , nucleic acid , folate , and steroid metabolism , as well as transport , signal transduction , development , behavior , protein translation , protein modification , protein folding , proteolysis , defense/immune response , muscle contraction , cytoskeleton , and exoskeleton . In comparison with a set of 143 predicted temperature-entrained circadian transcripts that is described in more detail below ( see Figure 4 and Materials and Methods ) , the temperature-driven transcripts show a higher frequency of functions associated with transport , transcription , translation , development , proteolysis , and protein folding , but a lower frequency of functions associated with circadian behavior and carbohydrate metabolism . All available microarray time courses from wild-type flies during and after temperature entrainment were used to define a set of temperature-entrained circadian transcripts . In order to be included , transcripts had to meet several noise filters and show a highly significant 24-h Fourier component as well as significant 24-h autocorrelation ( see Materials and Methods ) across the complete 2x CA/4x AA dataset . In order to avoid transcript oscillations that were merely temperature-driven , it was required that they also show a significant 24-h Fourier component and exceed background noise in an analysis of AA data only ( see Materials and Methods ) . The resulting set of 143 temperature-entrained circadian transcripts is presented in a phasegram ( Figure 4 ) , and the functions associated with these transcripts are described in Table S1 . As noted above , the set of temperature-entrained circadian transcripts shows some differences in its functional representation relative to the set of temperature-driven transcripts . Perhaps the most remarkable functional enrichment among temperature-entrained circadian transcripts is found for the takeout ( to ) gene family , which has been proposed to contribute to courtship behavior , starvation response , and olfaction [31 , 32] . Seven of the 21 members of this gene family show a strong temperature-entrained circadian expression component , whereas a robust temperature-driven response ( represented by oscillating expression in a temperature cycle in tim01 flies ) is only found for one gene , which happens to also show circadian regulation ( Table S1 ) . We also defined a set of 172 light-entrained circadian transcripts by applying similar selection criteria to the results from previous analyses of all available LD/DD microarray time course data [27] ( see Materials and Methods ) . The overlap between the two datasets ( 49 transcripts ) is highly significant and involves about a third of the transcripts in each set ( Figure 5A and Table S1 ) , which is considerably more than , for example , the overlap of either set with the set of 164 temperature-driven transcripts from Figure 4 ( overlap of 22 with the temperature-entrained and 13 with the light-entrained circadian transcripts ) . The fact that two-thirds of the predicted temperature-entrained circadian transcript profiles are not represented in a stringently selected set of light-entrained circadian transcript profiles does not mean that they do not show significant light-entrained circadian oscillations . The degree of overlap between the independent selections is quite sensitive to factors such as choice of selective cut-off criteria and reproducibility of the relative rankings of circadian transcripts . Nevertheless , the fact that the overlap is incomplete could indicate that there are both light- and temperature-specific circadian transcript oscillations . To examine the relative importance of such Zeitgeber-specific entrainment , we tested whether the overlap between two independent 4-d DD datasets was substantially larger than that between each of the two DD sets and our 4-d AA dataset; the results are illustrated in Figure 6 . The two independently detected DD circadian expression programs did not obviously share more transcripts with each other than with the AA circadian expression program . There is , therefore , no evidence for widespread temperature-specific or light-specific circadian expression . To further investigate this issue , six potential temperature-specific and three potential light-specific circadian transcript profiles were verified using northern blots . All six of the potential temperature-specific circadian transcripts show circadian oscillations in LD/DD ( Figure S1A ) , whereas none of the three transcripts with a predicted light-dependent circadian response exhibited clock-dependent regulation in response to temperature entrainment ( Figure S1B ) . These results are consistent with the notion that most , but not all , light-entrained circadian transcripts are also entrained by an environmental temperature cycle . Identification of a set of transcripts that show especially robust circadian regulation in response to both entrainment by light and entrainment by temperature allowed us to directly compare the circadian phases dictated by these two Zeitgebers . When all available LD/DD and CA/AA time course microarray data for this set is ordered in a phasegram according to the observed light-entrained phase , it becomes apparent that temperature-entrained phases are directly correlated with independently determined light-entrained phases , suggesting that mutual phase relationships among circadian transcripts may be preserved under entrainment of either light or temperature ( Figure 5A ) . When estimated light-entrained and temperature-entrained peak phases are directly plotted against each other , it becomes apparent that their relationship can be described by a linear function that represents a fixed phase offset ( Figure 5B ) . When the light-entrained phase is determined relative to lights-on and the temperature-entrained phase is determined relative to the onset of the thermophase , the value of the offset is ∼6 h . In the context of natural environmental conditions ( Figure 1 ) , however , the light/dark and temperature cycles are aligned differently , with the temperature minimum occurring just before dawn and the temperature maximum delayed relative to the time of maximum solar irradiance . It would be , therefore , more appropriate to assign temperature-entrained phases relative to the time of subjective dawn ( the middle of the cryophase ) . In this context , the observed temperature-entrained and light-entrained molecular phases would , in fact , roughly coincide , indicating cooperative entrainment by environmental cycles in light and temperature under most natural conditions . Convergence of photic and thermal entrainment is also observed at the behavioral level ( Figure S2 ) . During LD cycles , flies are preferentially active around the dark-to-light and light-to-dark transitions , with a siesta in the middle of the day . Although the sudden changes in environmental light at the transitions directly elicit behavioral startle responses in a clock-independent manner , a functional clock is required for control of a major circadian activity component at dusk and a minor circadian activity component at dawn ( e . g . , [18] ) . During temperature entrainment , the outlines of a similar pattern of clock-dependent activity can be recognized with onset of dawn-associated activity occurring during the last half of the cryophase and dusk-associated activity coinciding with the middle of the thermophase ( Figure S2 ) . We further examined the cooperative effects of light and temperature on entrainment of locomotor activity behavior . We entrained flies to an LD cycle and then released them into constant darkness in which a temperature cycle was given either “in phase” ( i . e . , light onset precedes onset of the thermophase by 6 h ) or “out of phase” ( i . e . , thermophase onset precedes light onset by 6 h ) . When light and temperature are given in phase , the average time of activity offset after 5 d of temperature is the same as the previous light phase ( Zeitgeiber time [ZT] 13 . 23 ± 1 . 32 versus ZT 13 . 98 ± 0 . 95; Figure 7 ) . However , when light and temperature are given out of phase , the average time of activity offset gradually advances over 5 d of temperature to approximately 10 h earlier than that measured in LD ( ZT 3 . 70 ± 2 . 90 versus ZT 13 . 46 ± 0 . 56; Figure 7B and 7C ) . Thus , the temperature- and light-entrained circadian behavioral phases show essentially the same relationship as the molecular phases that we discussed above . Under “natural” environmental conditions ( i . e . , when light onset and offset coincide with the middle of , respectively , the cryophase and thermophase ) both molecular and behavioral rhythms are entrained cooperatively by temperature and light . The well-characterized core clock transcripts that are known to oscillate under light entrainment ( per , tim , Clk , cry , vri , Pdp1ɛ ) continue to cycle under temperature entrainment in wild-type flies , while the noncycling core clock transcripts ( cyc , dbt ) remain constant ( Figure 8 ) . In LD cycles , the levels of the core clock transcripts remain at constitutive high or low levels in arrhythmic tim01 flies [4 , 30 , 33–35] . Analysis of the tim , per , Clk , and cry transcripts in arrhythmic tim01 flies exposed to a temperature cycle , however , reveals underlying temperature-driven responses that do not persist under constant conditions ( Figure S3 ) . These temperature-driven oscillations are mostly out of phase with the circadian transcript rhythms observed in wild-type flies ( Figure S3 ) . In order to maintain essentially the same circadian expression program for its core components in the presence or absence of an environmental temperature cycle , therefore , the clock has to largely neutralize the direct effects of temperature on their expression . The phase relationships between core clock transcripts ( i . e . , per , tim , vri , and Pdp1ɛ oscillate antiphase to Clk and cry ) are largely maintained in wild-type , temperature-entrained flies . Entrainment to temperature cycles , therefore , appears to promote the same overall clock organization and function as light entrainment , although with an important distinction that involves per and tim RNA expression . In light entrainment and subsequent free-run , per and tim transcription is tightly coupled at all times ( Figure S4 ) . In temperature entrainment , however , per and tim are uncoupled . This is due to both an advance in per expression and a delay in tim expression ( compared to free-run; Figure S4 ) . This divergence is absent in constant conditions following temperature entrainment . It is also absent at the protein level ( Figure S5 ) . One possible explanation for this discrepancy is that in temperature cycles a shift in the timing of per and tim RNA accumulation may be required to maintain coordinately phased accrual of the PER and TIM proteins . The advance in per expression could in theory be explained by the effects of a thermosensitive splicing event in the 3′ UTR of per , which is thought to enable flies to seasonally adapt to cold , short days [36–38] . It is important , however , to address this hypothesis experimentally , as alternative splicing of per has not been directly examined in the context of a thermocycle . The delay in tim expression is associated with thermosensitive splicing . A second tim transcript ( referred to as timcold ) is observed during temperature entrainment , especially during the cold phase ( Figure 9A ) . timcold is expressed at low levels in light entrainment at 25 °C ( Figure 9B ) , but in light entrainment performed at 18 °C it is the dominant isoform ( Figure 9C ) . Overall tim transcript levels appear to be increased at 18 °C relative to 25 °C by a factor of 1 . 5–2 ( unpublished data ) . While the canonical shorter transcript of tim oscillates with a phase that differs from that of per during temperature entrainment , timcold cycles in phase with per ( Figure S4 ) . Total tim transcript levels in the presence of the 18 °C /25 °C thermocycle follow the same pattern as the canonical transcript , albeit at a somewhat lower peak/trough ratio ( ∼3-fold versus ∼5-fold ) . Reverse transcriptase-PCR and northern blot analyses reveal that , in timcold , the last tim intron ( ∼850 bp ) is retained ( unpublished data ) . This unspliced form of tim has a premature stop codon that would putatively result in a protein about 3 . 5 kDa smaller than the full-length TIM protein . The missing fragment corresponds to a small piece of the cytoplasmic localization domain [39] . Western blots reveal the presence of two TIM isoforms at 18 °C in light entrainment , the lower of which is downregulated at 25 °C ( Figure 9D ) . Further work will be needed to understand the role of tim alternative splicing .
Transcriptional regulation of per and tim appears to be different in light and temperature entrainment . Whereas in light entrainment per and tim RNA expression is tightly coupled at all times , in 18 °C/25 °C temperature entrainment per RNA levels peak before tim RNA levels . This is a result of a temperature-induced advance in per expression and delay in the expression of the predominant tim transcript . Differences in per and tim regulation have been suggested based on the observation that these transcripts show different rates of degradation in response to a light pulse in the context of the long period mutant timul [40] . In addition , while at lower temperatures per expression is upregulated in LD and DD , tim has been reported to be downregulated in LD and barely oscillatory in DD [38] . Further , while the phases of both per and tim appeared advanced at lower temperatures , the advance in per was interpreted as a result of faster accumulation , while the advance in tim was thought to represent more rapid degradation [38] . It has also very recently been reported that tim , but not per , transcript levels are upregulated in response to light pulses at cold temperatures [41] . It is noteworthy , however , that the probe used in several previous studies [24 , 38 , 41] to evaluate tim transcript expression with RNase protection assays may not have efficiently detected the timcold isoform since it spans the exons flanking the intron maintained in timcold . This issue is illustrated by quantitation of the data in Figure 9C , which confirms the predicted decrease of tim transcript in the first day of DD at 18 °C [38] for the predominant isoform , but not for timcold , which shows a more prominent peak in expression ( unpublished data ) . Additional analyses that take into account the contribution of the timcold isoform will , therefore , be needed to complement previous studies in order to more fully explore tim transcript responses . One of the factors involved in the reported differential expression of per and tim may be the alternative splicing of both transcripts . Much of the recent molecular work on temperature and the circadian clock has focused on the alternative splicing of an 89-bp intron in the 3′ UTR of per , an event thought to be important in seasonal adaptation [36–38] . Short , cold days lead to increased amounts of the spliced per variant , resulting in an earlier increase in PER protein abundance and an advanced phase of locomotor activity . Warmer temperatures result in less of the spliced variant , especially during the day . This appears to be a clock-dependent effect that results in the fly moving its behavior to the later ( cooler ) part of the day . Thus , per splicing allows the fly to adapt to changes in both temperature and photoperiod by regulating the amount of available PER protein . per alternative splicing is thought to be important in seasonal adaptation , as long photoperiods counteract the cold-induced behavioral advances by delaying the accumulation of TIM , in turn rendering prematurely produced PER unstable . Thus , the fly is able to integrate information from both light and temperature to generate behavior that is aligned to the environmental day . Regulation of per splicing in the presence of an environmental temperature cycle as compared to constant temperature needs to be investigated . Temperature-dependent alternative splicing of tim is described here . At 18 °C , the last intron of tim is preferentially retained , resulting in a premature stop codon and a truncated protein . Although the expression of the predominant tim transcript is delayed relative to per , timcold cycles in phase with per . The differential expression of the two tim transcripts could reflect temperature-dependent control of splicing or of the stability of one of the splice forms . We are still in the process of ascertaining the functional significance of the production of timcold transcript . It does , however , appear that the alternative splicing and differential regulation of per and tim are responsible for finely tuning the clock in response to changing environmental conditions , thus adding an additional level of complexity to the clock . Different groups of clock-bearing cells in the fly have been shown to regulate different rhythmic processes . For example , locomotor activity and eclosion rhythms , arguably the best-characterized rhythmic behaviors in Drosophila , require the ventral lateral neurons ( LNvs ) and the neuropeptide , Pigment Dispersing Factor [42–44] . Cyclic olfactory responses do not depend on the LNs or Pigment Dispersing Factor , but instead depend on the antennal neurons [45 , 46] . Egg-laying rhythms also appear to be regulated independently of the LNvs and Pigment Dispersing Factor [47] . Thus , the image of the circadian clock as a single entity is transforming into a more complex model . A system of two coupled oscillators was proposed for the Drosophila clock almost 50 y ago [48] . In this model , the master or A oscillator is autonomous , light-sensitive , and temperature-compensated . The slave or B oscillator , which is coupled to and driven by A , is responsive to temperature but not light . The evidence for this two-oscillator model came from the different responses in eclosion rhythms to light and temperature . Whereas light pulses administered at different times of day resulted in steady-state phase advances or delays , the phase changes resulting from temperature pulses were transient . The researchers concluded that the steady-state phase changes in response to light were a result of the eventual realignment of the A oscillator to the light signal . The transient responses to temperature pulses were proposed to be a result of temporary temperature-induced disturbances in B , with the return to the previous phase reflecting the A oscillator's resumption of control over B . Coupled oscillators have been proposed in a variety of organisms [49] . For example , different genes are expressed with different period lengths in some Synechoccus mutant backgrounds [50] , and bioluminescence rhythms in Gonyaulax have been shown to be regulated by two oscillators that respond to different wavelengths of light [51] . In Neurospora , strains carrying null alleles of frequency ( frq ) , white collar-1 ( wc-1 ) , or white collar-2 ( wc-2 ) still show a conidial banding rhythm . Although the “FRQ-less oscillators” [52] responsible for these rhythms have lost most characteristics of a circadian clock , they can be entrained by Zeitgebers such as temperature cycles [53 , 54] , rhythms in nitrate reductase activity [55] , and transfer from light to dark [56] . This suggests that FRQ-less oscillators are slave oscillators , requiring the master FRQ-dependent clock to produce stable circadian rhythms , yet with the ability to function independently of the master clock in a non-circadian manner [57] . A system of coupled oscillators has recently been demonstrated in the regulation of the morning and evening peaks of locomotor activity in the fly [58–60] . The morning oscillator requires the presence of the LNvs , while the evening oscillator requires the dorsal lateral neurons . It was further shown that the evening oscillator is set by the morning oscillator by generating flies in which the morning and evening oscillators have different free-running periods [60] . However , despite the parallels to Pittendrigh's original model , there is no published evidence that these or other oscillators would differentially respond to temperature , as opposed to light , as a Zeitgeber . So while it appears there is a multicellular clock network in Drosophila that is reflected by coordinate yet independently regulated outputs , the data presented here suggest that the response to multiple inputs , such as light and temperature , would still be integrated by a single autonomous clock mechanism ( Figure 10 ) . In today's jargon we would describe Pittendrigh's B oscillator as a circadian output pathway that can show direct clock-independent responses to temperature . The following observations support the hypothesis of a single , integrative transcriptional oscillator . First , the same set of core clock components ( including PER , TIM , CLK , and CYC ) appears to be required for producing both light-entrained and temperature-entrained oscillations [18 , 19] . The global transcriptional signatures of arrhythmic tim01 flies that we found after thermocycle treatment resemble those found after photocycle treatment [27] and do not exhibit obvious circadian rhythms . In addition , our results confirm the absence of circadian oscillations for core clock gene transcripts in the tim01 fly heads . Second , it is likely that the set of transcripts entrainable by thermocycles is closely related to the set of transcripts entrainable by light . Although we cannot formally exclude the existence of circadian rhythms that specifically require temperature entrainment , we have found none so far . Third , the phases of the transcripts that oscillate in response to both photo- and thermocycles maintain the same mutual phase relationships after entrainment by light or temperature . The phase observed at the onset of the thermophase is systematically advanced by about 6 h relative to the phase at the onset of light . Given the size of the delay that is commonly found between the environmental profiles for temperature relative to that of daylight , these results indicate cooperative entrainment by light and temperature under common natural circumstances . A response to temperature would be well integrated with the expected light cycle were it also supplied , and vice versa . Fourth , the temperature- and light-entrained phases of PER and TIM protein expression [13] ( see also Figure S5 ) reflect the same relationship that we observed for the genome-wide circadian transcript signatures . This observation is consistent with the hypothesis that both light and temperature act via the same PER/TIM-dependent oscillator to generate circadian transcript profiles . Fifth , the entrained phase of locomotor activity behavior appears to follow the molecular circadian phase observed in temperature or light entrainment . Our ability to accurately predict the phase of clock neuron–controlled circadian locomotor behavior based on our analysis of circadian transcript rhythms in a preparation of whole heads , which mostly represents peripheral clock cells , suggests that temperature entrainment just as light entrainment produces similar phases in peripheral clock cells and clock neurons . This result can be verified and extended in a future study by direct examination of the temperature-entrained molecular phase in the various subsets of clock neurons . In summary , our analyses revealed that thermocycle entrainment and photocycle entrainment produce very similar circadian expression profiles in fly heads , and that under common natural conditions light and temperature are expected to entrain both molecular and behavioral circadian rhythms cooperatively . As pointed out above , our results are in agreement with the notion that a single transcriptional clock is responsible for producing all light-entrained and temperature-entrained circadian rhythms . Nevertheless , we cannot formally exclude the existence of a specialized temperature-entrained oscillator that is coupled to the general transcriptional clock circuits . Such a theoretical temperature-entrained oscillator could have eluded detection in our analyses if it was located outside the head or in a small subset of the cells in the head or if it produced non-transcriptional circadian signals . Elucidation of the mechanisms of thermocycle entrainment will constitute an important next step in defining the temperature-entrained circadian oscillator ( s ) .
The wild-type strains used were yellow white ( y w ) , Canton S , and cinnabar brown ( cn bw ) ; in addition , comparative analyses were performed for y w; tim01 arrhythmic mutant flies . The flies were raised on standard yeast cornmeal medium . For light experiments ( described previously; see [27 , 28] ) , adult flies were entrained to 12 h of light followed by 12 h of darkness ( LD , where ZT0 is lights-on and ZT12 is lights-off ) at 25 °C for 5 d and subsequently released into constant darkness ( DD , where circadian time [CT]0 is subjective lights-off ) . They were harvested onto dry ice every 4 h ( unless specified otherwise ) during the last day of entrainment and for 1–2 d of DD . Temperature experiments were conducted entirely in the dark except for the initial seeding of parental bottles . Adults ( parental ) were placed in fresh media and allowed to mate and lay eggs at 25 °C for 5 d . The parents were cleared and the next generation was kept at 25 °C until early pupal stages . Next , they were transferred to a temperature cycle of 12 h of 18 °C followed by 12 h of 25 °C ( CA , where CA0 is onset of 18 °C and CA12 is onset of 25 °C ) until eclosion occurred . The newly eclosed flies were transferred to fresh media and allowed to entrain for an additional 4 d and then released into a constant temperature of 25 °C ( AA ) . They were harvested onto dry ice every 4 h during the last day of entrainment and for one to two and a half days in AA ( as specified: Day 1 = AA2–22 , Day 2 = AA26–46 , partial Day 3 = AA50–58 ) . The frozen flies were vortexed and passed through a series of sieves in order to isolate the heads for RNA or protein extraction . Individual flies were monitored and their locomotor activity was analyzed with the Drosophila Activity Monitoring System IV ( TriKinetics , http://www . trikinetics . com ) . For LD/DD experiments , the flies were reared under standard lighting conditions and monitored at 25 °C in LD and/or DD as specified . Temperature experiments were conducted entirely in the dark ( unless otherwise specified ) . The flies were raised at 25 °C until the pupal stage and were then entrained to CA until eclosion . Individual flies were monitored and their locomotor activity analyzed as above both during entrainment ( CA ) and free-run ( AA ) . Period lengths were calculated using ClockLab Software ( ActiMetrics , http://www . actimetrics . com ) . Total RNA was extracted from approximately 100 μl of adult heads per time point using either RNA-STAT60 ( Tel-Test , Incorporated , http://www . tel-test . com ) or guanidinium thiocyanate followed by centrifugation in Cesium chloride solution . 15–30 μg of total RNA was denatured for 5 min at 65 °C and resolved on a 1% formaldehyde-agarose gel ( 20 mM MOPS [pH7] , 5 mM NaOAc , 1 mM EDTA ) . The resolved RNA was transferred to Nytran membrane ( Schleicher & Schuell , http://www . arraying . com ) in 10x SSC overnight . Probe templates were radiolabeled as specified for the DECAprime II kit ( Ambion , http://www . ambion . com ) . Hybridizations were carried out at 55 °C in UltraHyb solution ( Ambion ) supplemented with denatured fish sperm DNA ( Roche , http://www . roche . com ) . Radioactive signals on the blots were visualized and quantitated with either a Storm or Typhoon Phosphorimager ( Molecular Dynamics ) and the results plotted in Microsoft Excel ( http://www . microsoft . com ) . Fourier analyses were also performed on the northern data as indicated . Total protein was extracted from about 35 μl of adult heads per time point in 75 μl Head Extraction Buffer ( 100 mM KCl , 20 mM Hepes [pH 7 . 5] , 10% glycerol , 10 mM EDTA [pH 8] , 0 . 1% Triton X-10 , 50 mM NaF , 1 mM DTT ) with 1x protease and phosphatase inhibitors ( Roche ) using a handheld homogenizer ( Kontes , http://www . kontes . com ) . Samples were centrifuged at 14 , 000 rpm for 15 min at 4 °C . The supernatant was transferred to a new tube and centrifuged as above for an additional 10 min . 15–30 μg of total protein were resolved on 6% SDS polyacrylamide gels and transferred to nitrocellulose membrane ( Schleicher & Schuell ) . Membranes were blocked for at least 1 h at room temperature with 5% nonfat dry milk in 1x TBST . Primary antibodies were diluted in blocking solution ( 1:10 , 000 for α-PER [rabbit] , 1:2 , 000 for α-TIM [rat] ) and incubated with the membranes at 4 °C overnight . The membranes were washed three times for 10 min each in 1xTBST and incubated with secondary antibodies ( 1:10 , 000 ) ( Jackson ImmunoResearch , http://www . jacksonimmuno . com ) for 1 h at room temperature . The membranes were washed as before and detection was carried out using ECL ( Amersham Pharmacia Biotech , http://www . gehealthcare . com ) . RNA was extracted with RNA-STAT60 ( Tel-Test , Incorporated ) in the same manner as for northern blots . cDNA was generated using the ThermoScript RT-PCR System ( Invitrogen , http://www . invitrogen . com ) as described by the manufacturer with one exception: cDNA synthesis was carried out at 50 °C for 90 min with Oligo ( dT ) 20 . cDNA ( 3 μl ) was used in subsequent PCR reactions with AccuPrime Pfx DNA Polymerase ( Invitrogen ) . Two sets of y w flies and one set of arrhythmic mutant y w; tim01 flines were collected as described above for 1 d in CA and 1 d in AA . An additional set of cn bw flies was collected for the first 2 d in AA . RNA was extracted from adult heads with guanidinium thiocyanate followed by centrifugation in Cesium chloride solution . 50 μg of the RNA was further purified over RNeasy columns ( Qiagen , http://www1 . qiagen . com ) according to instructions from the manufacturer . 25 μg of the purified RNA was used to generate biotin-labeled cRNA probe as described in the Affymetrix GeneChip manual . T7-d24 primers ( MWG Biotech , http://www . mwg-biotech . com ) , Superscript Choice ( Life Technologies , Invitrogen ) , and enzymes from New England Biolabs ( http://www . neb . com ) were used to synthesize cDNA . The ENZO Bioarray High Yield RNA transcript labeling kit was used for in vitro transcription reactions . Hybridization , washing , staining , and scanning of the target cRNA to the Affymetrix Drosophila Genome 1 arrays were carried out according to the Affymetrix GeneChip manual . The robust multi-array average ( RMA ) single algorithm [61] was used to prepare microarray data from each experiment . Using Fourier analysis , 24-h spectral power ( F24 ) was calculated for appended time course experiments as in [27] . To estimate probability values and FDR , a permutated background model was used , in which time ordering of the real data was permuted 1 , 000 times to give a background distribution of F24 . These were divided into a number of quantiles equal to the original number of probe sets and compared to F24 from the real data . Two types of probability values were calculated in association with F24: local p-values assigned to each probe set represent the odds of observing a F24 score of equal or higher strength after random permutation of the time order for the data series of that probe set , whereas global p-values represent the odds of observing an equal or higher F24 in the distribution calculated after random permutation of the time order for the data series of all of the probe sets . The local p-values are used in the selections performed for Figures 3 , 4 , and 5 , whereas the global p-values are used to describe the 24-h periodic expression programs as represented in the graphs of Figures 2 and 6 . For the analyses in Figures 2 and 6 , only probe sets for which more than half of the values exceeded a 20th percentile cut-off were considered . In Figure 2B–2E , the number of selected rhythmic transcript profiles is plotted as a function of the threshold that is applied to the global p-value . The FDR in Figure 2A simply represents the ratio of the number of transcript profiles selected from the randomly permuted dataset over the number of transcript profiles selected from the real data . The LD data used in Figures 2 and 6 represent the y w ( 1 ) and y w ( 3 ) time course datasets described in references [27 , 28] , whereas the DD data in Figure 2 represent these time courses plus 2 d of data from [29] formatted as described previously in [27]; the DD1 dataset in Figure 6 is identical to the DD data in Figure 2 , whereas the DD2 dataset consists of the cn bw time course described in [28] plus 2 d of data from [62] and the average for the day of data from [63] . To select the 164 temperature-driven transcripts illustrated in Figure 3 and Table S1 , the local p-value for the CA ( 2x wt + 1x tim01 ) data had to be <0 . 05 , the average absolute daily range of RMA expression values had to be >0 . 3 for both the CA ( 2x wt ) and the CA ( 2x wt + 1x tim01 ) data and the p-value found for a Kruskal-Wallis test of significant variation with daily time across the CA ( 2x wt + 1x tim01 ) data had to be <0 . 025 . To select the 143 temperature-entrained circadian transcripts illustrated in Figure 4 and Table S1 , the local p-value for the CA/AA ( 2x/4x wt ) data had to be <0 . 001 , the average absolute daily range of RMA expression values had to be >0 . 3 , and the p-value found for a Kruskal-Wallis test of significant variation with daily time across the CA/AA ( 2x/4x wt ) data had to be <0 . 01 , and , to avoid light-driven effects , the local p-value for the AA ( 4x wt ) data had to be <0 . 05 . The set of core circadian transcripts in Figure 5 and Table S1 was defined by the overlap of the 143 temperature-entrained circadian transcripts and a set of 172 light-entrained circadian transcripts that was defined based on the following criteria: local p-value for LD/DD ( 8x/9x wt ) had to be <0 . 001 [27] , the average absolute daily range of RMA expression values had to be >0 . 3 , and the p-value found for a Kruskal-Wallis test of significant variation with daily time across the LD/DD ( 8x/9x wt ) data had to be <0 . 05 , and , to avoid light-driven effects , the local p-value for the DD ( 9x wt ) data had to be <0 . 05 . Analysis results are made available at http://biorhythm . rockefeller . edu .
The microarray data generated for this study have been deposited under accession number GSE6542 in the Gene Expression Omnibus data repository ( http://www . ncbi . nlm . nih . gov/geo ) . FlyBase ( http://www . flybase . net ) accession numbers for the Drosophila genes discussed in this paper are Clk ( FBgn0023076 ) , cry ( FBgn0025680 ) , Pdf ( FBgn0023178 ) , Pdp1 ( FBgn0016694 ) , per ( FBgn0003068 ) , tim ( FBgn0014396 ) , and vri ( FBgn0016076 ) . | A key adaptation to life on Earth is provided by internal daily time-keeping mechanisms that allow anticipation of the alternations between night and day . To act as reliable time-keeping mechanisms , circadian clocks have to be able to synchronize to environmental time cues , maintain ∼24-h rhythms under constant conditions , run at approximately the same pace over a range of environmental temperatures , and efficiently communicate time-of-day information to other biological systems . Clock-controlled oscillations in gene expression play an essential role in producing overt circadian rhythms . For most organisms , light/dark cycles appear to constitute the most powerful entrainment cue , but daily temperature cycles have also been demonstrated to efficiently synchronize circadian rhythms . This study uses the fruit fly Drosophila melanogaster as a model to compare the clock-dependent and clock-independent daily gene expression rhythms produced in response to light/dark cycles versus temperature cycles . A broad temperature-driven expression program was found in the heads of both wild-type and arrhythmic mutant flies , but wild-type flies also exhibited a more specific temperature-entrained circadian expression response that resembled the circadian response following light entrainment . The phase relationship between the temperature- and light-entrained circadian rhythms suggests that in nature light and temperature act cooperatively to synchronize the circadian clock . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"biochemistry",
"none",
"cell",
"biology",
"neuroscience",
"drosophila",
"genetics",
"and",
"genomics"
] | 2007 | Integration of Light and Temperature in the Regulation of Circadian Gene Expression in Drosophila |
Invariant natural killer T ( iNKT ) cells are an evolutionary conserved T cell population characterized by features of both the innate and adaptive immune response . Studies have shown that iNKT cells are required for protective responses to Gram-positive pathogens such as Streptococcus pneumoniae , and that these cells recognize bacterial diacylglycerol antigens presented by CD1d , a non-classical antigen-presenting molecule . The combination of a lipid backbone containing an unusual fatty acid , vaccenic acid , as well as a glucose sugar that is weaker or not stimulatory when linked to other lipids , is required for iNKT cell stimulation by these antigens . Here we have carried out structural and biophysical studies that illuminate the reasons for the stringent requirement for this unique combination . The data indicate that vaccenic acid bound to the CD1d groove orients the protruding glucose sugar for TCR recognition , and it allows for an additional hydrogen bond of the glucose with CD1d when in complex with the TCR . Furthermore , TCR binding causes an induced fit in both the sugar and CD1d , and we have identified the CD1d amino acids important for iNKT TCR recognition and the stability of the ternary complex . The studies show also how hydrogen bonds formed by the glucose sugar can account for the distinct binding kinetics of the TCR for this CD1d-glycolipid complex . Therefore , our studies illuminate the mechanism of glycolipid recognition for antigens from important pathogens .
Invariant NKT cells ( iNKT ) are an evolutionarily conserved population of T lymphocytes able to respond to lipid antigens when presented by CD1d , a non-classical MHC class I–like molecule [1] . Antigen recognition by iNKT cells is mediated by a semi-invariant αβ T cell receptor ( TCR ) formed by a conserved Vα14-Jα18 rearrangement ( Vα24-Vα18 in humans ) , and a limited panel of pairing β chains ( Vβ8 . 2 , Vβ7 , Vβ2 in mouse; Vβ11 in humans ) . The number of antigens recognized by these cells has increased in the last few years , following the discovery of α-galactosylceramide ( α-GalCer ) , the prototypical iNKT antigen [2] . Although the nature of the predominant self-antigens recognized by iNKT cells still remains controversial , important progress has been made in describing the microbial antigens recognized by this cell type . Glycosphingolipids from Sphingomonas spp . and diacylglycerol ( DAG ) ligands from Borrelia burgdorferi , the causative agent of Lyme disease , were identified to stimulate iNKT cells in a CD1d/TCR-dependent manner [3]–[7] . As Sphingomonas spp . and B . burgdorferi are not responsible for widespread or lethal diseases , we considered it possible that more pathogenic organisms also express iNKT antigens , which would account for the highly conserved nature of the CD1d-iNKT TCR interaction . Indeed , recent studies identified the structures of DAG compounds from the highly pathogenic Streptococcus pneumoniae ( S . pneumoniae ) and Group B streptococcus ( GBS ) , which were able to stimulate iNKT cells [8] . In vitro and in vivo assays demonstrated surprisingly strict requirements for these antigens in activating iNKT cells . The most potent S . pneumoniae antigen , Glc-DAG-s2 , is characterized by having a sn-3 linked glucose , a sn-1 linked palmitic acid ( C16∶0 ) , and most importantly , the presence of cis-vaccenic acid ( C18∶1 , n-7 ) in position sn-2 of the glycerol moiety . This uncommon fatty acid was required for significant activity , since the positional isomer with a vaccenic acid in position sn-1 failed to elicit a strong activation , as did the homologous compounds containing an oleic acid ( C18∶1 , n-9 ) . Moreover , the same antigen showed the ability to stimulate both mouse and human iNKT cells , unlike the previously characterized DAG antigens from B . burgdorferi [6] . Interestingly , previous studies showed that glucose-containing glycolipids are relatively weaker antigens compared to the one containing galactose or galacturonic acid [2] , [3] , [9] , [10] , while the glucose isomer of the B . burgdorferi glycolipid 2c ( BbGL-2c ) is not antigenic at all [6] . It is therefore surprising that the glucose-containing S . pneumoniae Glc-DAG-s2 is such a potent antigen in eliciting iNKT cell responses . In order to determine the molecular basis for the stringent structural requirements for recognition of the S . pneumoniae antigen Glc-DAG-s2 , and to further analyze the mechanism of the mouse CD1d ( mCD1d ) -iNKT TCR complex formation , we determined the structure of the mCD1d-Glc-DAG-s2-iNKT TCR complex by X-ray crystallography and we analyzed the role of the F′ roof in the formation and stability of mCD1d-iNKT TCR complexes . Our data show how the combination of cis-vaccenic acid and glucose is required for the formation of novel protein-antigen contacts , resulting in the relatively strong affinity of this ligand for the iNKT TCR .
Previous biophysical analysis of the kinetics of interaction of the mCD1d-Glc-DAG-s2 complex with the iNKT TCR revealed a TCR interaction with a low micromolar affinity [8] . Overall , Glc-DAG-s2 complexes with mouse CD1d are characterized by a comparable but slightly higher affinity for the iNKT TCR compared to those containing the other known bacterial DAG antigen BbGL-2c ( KD of 4 . 4 and 6 . 2 µM , respectively ) , consistent with the similar antigenic potencies of these compounds [8] , [11] . However , complexes containing Glc-DAG-s2 have significantly different binding kinetics compared to those containing BbGL-2c , with Glc-DAG-s2 showing considerably slower association and dissociation rates . In order to investigate the molecular basis for this different kinetic behavior and the role of the unique structural features of this ligand in determining its antigenicity , we determined the crystal structure of the mCD1d-Glc-DAG-s2-iNKT TCR complex at 2 . 7 Å resolution ( Figure 1 , Table 1 ) . The structure shows the conserved “parallel" docking mode of the iNKT TCR on the CD1d-ligand complex ( Figure1A ) [12]–[14] . As a consequence of this unique binding mode , the TCR α chain mediates the majority of the contacts with the CD1d-Glc-DAG-s2 complex , with additional contacts with CD1d provided by the CDR2β , CDR3β , and , to a lesser extent , the CDR1β loops ( Table S1 ) . Well-defined , unbiased density was present for the ligand , superior to what has been observed for the mCD1d-Glc-DAG-s2 complex in absence of the TCR [8] , suggesting that the ligand adopts a more rigid and ordered conformation upon TCR binding ( Figure 1B , Figure S1 ) . Similar to the antigens previously characterized , the TCR CDR1α and CDR3α loops exclusively mediate contacts between the TCR and the antigen ( Figure 2A ) . In particular , the TCR recognizes the 2′-OH and 3′-OH positions of the hexose ring via H bonds with Gly96 and Asn30 on the α chain , highlighting the importance of these two hydroxyl groups on the antigen in the formation of the complex . However , due to the presence of a glucose on Glc-DAG-s2 , the 4′ hydroxyl group is no longer able to interact with Asn30 on the α chain , in contrast to other galactose-containing glycolipids . Previous studies showed that the contacts between the ligand and the iNKT dominate the initial association phase of the interaction [9] . The loss of an H bond at the ligand-TCR interface , although not sufficient to abolish the binding of the iNKT TCR to the mCD1d-Glc-DAG-s2 complex , is therefore likely to decrease the association rate . When the structures of the mCD1d-Glc-DAG-s2 complex in the presence or absence of the TCR are compared , important conformational changes are observed for the ligand ( Figure 2B ) . Consistent with what was observed for the mCD1d-Glc-DAG-s2 structure in the absence of the TCR [8] , the sn-2 vaccenic acid is bound in the A′ pocket while the sn-1 palmitic acid occupies the F′ pocket . However , while in the absence of the TCR the vaccenic acid encircles the A′ pole in a clockwise manner , the opposite orientation is preferred in the ternary complex , although residual density also suggests some equilibrium between the two orientations . Moreover , upon TCR ligation , the glucose moiety is shifted by about 30 degrees clockwise around its glycosidic bond to assume a position at the center of the binding groove as observed for other TCR-bound glycolipid antigens ( Figure 2B ) . Similar to what has been described for BbGL-2c , this conformational change requires the breaking of several contacts with CD1d and the formation of new hydrogen bonds with the α2 helix of CD1d and the TCR α chain . In particular , a hydrogen bond with Arg79 on the α1 helix is lost while new polar contacts are formed with Asp153 and Thr156 on the α2 helix upon TCR binding , resulting in a final orientation conserved among α-linked sugars [13]–[15] . As proposed for BbGL-2c , these conformational changes likely contribute to the slower association rate of the TCR when binding DAG microbial antigen-mCD1d complexes compared to sphingolipid-containing antigens [14] . However , the presence of glucose on Glc-DAG-s2 results in an additional H bond between the antigen and the backbone of the α2 helix of CD1d , involving the carbonyl group of glycine 155 ( 3 . 2 Å , Figure 2B ) . Because this contact stabilizes a favorable binding conformation of the antigen in the binding groove , it is likely that this feature is contributing to the slower complex dissociation observed for this ligand compared to BbGL-2c , as the TCR has to invest less energy to lock the glucose into place . Moreover , comparison of the mCD1d-TCR molecular contacts in the two DAG antigens ternary complexes revealed a slightly optimized interface for Glc-DAG-s2 compared to BbGL-2c ( involving in particular additional salt bridges between CDR3α and CDR2β residues with mCD1d; Table S1; [14] ) , which could have a further stabilizing effect on the dissociation rate of the ternary complex . When the structures of the ternary complexes of the DAG antigens Glc-DAG-s2 and BbGL-2c are compared ( Figure 3 ) , it is interesting to note how the unsaturations present on the respective vaccenic and oleic acids are localized in the same portion of the mCD1d A′ pocket , suggesting a preference for this region of the groove for binding unsaturated alkyl chains . Consistent with this , it was previously noted that the presence of unsaturations improved the stability of the mCD1d-glycolipid complex , possibly due to the kink introduced in the alkyl chain by the unsaturated bonds , which could nicely sit at the bottom of the channel connecting the A′ pocket with the protein surface [16] , [17] . Assuming that an unsaturated fatty acid would preferentially bind in the A′ pocket , in the positional isomer of Glc-DAG-s2 having the vaccenic acid at the sn-1 position , this would result in a reversed orientation of the glycerol backbone in the binding groove . The reversed glycerol orientation would cause an unfavorable positioning of the glucose head , therefore explaining the lack of antigenic activity for this compound [8] . A superposition of the mCD1d-Glc-DAG-s2 and mCD1d-BbGL-2c complexes in presence of the TCR shows how the two hexose groups are oriented slightly differently at the opening of the binding groove ( Figure 3A ) . The lack of activity of the Glc-DAG-s2 analog containing an oleic acid in place of vaccenic acid suggests that the vaccenic acid is required for a more favorable orientation of the exposed glucose , possibly enhancing the ability of the glucose moiety to contact Gly155 and therefore positioning it in a more stable fashion in the correct orientation for TCR recognition . Galactose and glucose differ only with regard to the orientation of the 4′ hydroxyl on the hexose sugar ring , with the axial orientation for galactose and the equatorial orientation , i . e . , closer to the plane of the ring , for glucose . Despite the structural similarity of the two sugars , intriguingly , the galactose-containing version of the S . pneumoniae DAG glycolipid antigen with a sn-2 vaccenic acid , called Gal-DAG-s2 , was not able to activate mouse iNKT cell hybridomas ( Figure 4 ) . A drastically reduced response was also observed for the Gal-DAG-s1 ligand . This indicates that the presence of vaccenic acid in the A′ pocket of the mCD1d binding groove does not automatically confer antigenicity . When the stereochemistry of the 4′ carbon is inverted , converting the glucose of Glc-DAG-s2 into galactose , it becomes clear that the 4′-OH group will be too distant ( 4 . 3 Å , Figure S2 ) to engage Asn30 on the CDR3α loop , while at the same time losing the contact with Gly155 on the α2 helix of mCD1d . Even if a further reorientation of the galactose sugar by the iNKT TCR were possible , this would require an additional energetic toll , suggesting a rationale for the reduced activity of the Gal-DAG-s2 compound . It is therefore evident that the combination of the uncommon vaccenic acid and the glucose sugar , which is relatively weak in the context of other DAG antigens and glycosphingolipids , is required for the optimal positioning of the Glc-DAG-s2 antigen in the mCD1d binding groove for TCR recognition . The structures of the iNKT cell TCR in complex with different mCD1d-microbial antigens complexes showed how the iNKT TCR is able to induce conformational changes in both the ligand and mCD1d upon complex formation [14] . In particular , the insertion of amino acid Leu99α , located on the CDR3α loop of the iNKT TCR , between residues Leu84 , Val149 , and Leu150 above the F′ pocket of mCD1d , resulted in several new van der Waals ( VdW ) contacts and the formation of a hydrophobic surface above the pocket ( F′ roof ) . Consistent with this , a comparison of mCD1d-Glc-DAG-s2 structures before and after TCR binding reveals that the F′ roof also is formed for this antigen upon TCR binding ( Figure 5A ) . In order to understand and validate the role of the F′ roof in the formation of a more stable CD1d-Glc-DAG-s2-TCR complex , we mutated selected residues involved in the formation of the roof and characterized the ability of the mutated mCD1d proteins to stimulate iNKT cell hybridomas . As a complete removal of the F′ roof would likely result in an abrogation of binding , as demonstrated by the loss of function mutation of L99α in the TCR to alanine [18] , we chose mutations of the relevant position in mCD1d that would maintain the hydrophobic nature of their side chains , in order to perturb the F′ roof area without making a too drastic change . We therefore used site-directed mutagenesis to generate the following mCD1d substitutions: Leu84Val , Leu84Phe , the latter mimicking the human homolog , Val149Leu and Leu150Val , together with two control mutants , Met69Ala and Met162Ala from the area above the A′ pocket ( Figure 5A ) . Interestingly , we obtained drastically reduced expression yields for the Leu84Val mutant , and this construct was not tested further . The iNKT cell hybridomas Hy2C12 ( bearing the Vα14Vβ8 . 2 TCR used in our structural studies ) , Hy1 . 2 ( also Vα14Vβ8 . 2 ) , and Hy1 . 4 ( expressing a less common Vα14Vβ10 TCR ) were tested for their ability to respond to mCD1d-glycolipid complexes in an antigen presenting cell-free assay using mCD1d-coated plates . IL-2 secretion provided a measure of TCR stimulation . We stimulated the cells with either Glc-DAG-s2 or α-GalCer , the prototypical iNKT cell antigen that induces a preformed F′ roof on mCD1d [19] . When loaded with Glc-DAG-s2 or α-GalCer , all the mutants showed a reduced ability to stimulate the hybridoma ( Figure 5 ) . In particular , the mutants Leu84Phe and Val149Leu abrogated iNKT cell activation , while a slightly reduced activity was observed with the Leu150Val mutant . As the reduced response could be the consequence of impaired loading of these antigens on mCD1d , we measured the loading efficiency of α-GalCer on each mutant by surface plasmon resonance ( SPR ) using a monoclonal antibody ( L363 [20] ) specifically reactive to complexes of mCD1d with α-GalCer and analogs ( Figure 5B ) . Although the Leu84Phe and Met62Ala mutants showed lower levels of antigen loading compared to wild type mCD1d , loading on the mutated mCD1d proteins was never below 65% of the wild type control , and does not appear to correlate directly with the ability of the mutated proteins to stimulate the iNKT cell hybridoma . Although the lack of an antibody able to recognize the mCD1d-Glc-DAG-s2 complex did not allow us to assess the loading of this antigen onto the mCD1d mutants , we believe it is unlikely that the two ligands have radically different loading efficiencies , suggesting a critical role of the area above the F′ pocket in the TCR interaction with mCD1d/Glc-DAG-s2 and mCD1d/α-GalCer . We previously hypothesized that the formation of the F′ roof on CD1d affects the stability of the CD1d interaction with the iNKT TCR [14] . To validate this hypothesis we measured the effect of the F′ roof mutants on the binding kinetics of the mCD1d-iNKT TCR complex ( Figure 6 ) . Because of the relatively weak antigenicity of Glc-DAG-s2 compared to α-GalCer , the latter was chosen for SPR analysis . A comparison of the affinities shows how all the F′ roof mutants have weaker affinities for the iNKT TCR , Leu84Phe being the weakest with a ∼4-fold reduction compared to the wild type protein ( Figure 6 ) while the two control mutants showed affinities and kinetics similar to the wild type protein . Strikingly , the differences in affinities between mCD1d F′ roof variants derive mainly from faster dissociation rates for the mutant complexes , while the association rates are minimally affected by perturbation of the F′ roof . Mutation of residues Val149 and L150 appear less disruptive than mutation of Leu84 , as the Leu84Phe has the fastest dissociation rate . Taken together , these data suggest that the F′ roof is critical in determining the dissociation rate and therefore the stability of the mCD1d-TCR complex .
Activation of iNKT cells can result as a consequence of TCR-independent , IL12-dependent signals and/or through the recognition of self and foreign antigens by its semi-invariant TCR , with the latter mechanism playing an important role in modulating the overall response [21] . The recent discovery that highly pathogenic Gram-positive bacteria express antigens recognized by both mouse and human iNKT cells [8] therefore opens important perspectives for the development of therapeutic agents against pneumonia and meningitis , while also suggesting a potential rationale for the conserved features of the CD1d-TCR interaction among different mammalian species . Surprisingly , the S . pneumoniae Glc-DAG-s2 antigen presents unusual chemical features in both its lipidic and polar portions when compared to the previously characterized iNKT cell microbial antigens from Sphingomonas spp . or Borrelia burgdorferi . Instead of the α-galactose or α-galacturonic acid found on these antigens , and on the prototypical antigen α-GalCer , the otherwise weaker α-glucose is found in S . pneumoniae as well as in another gram-positive pathogen , GBS . Furthermore , the sugar is α-linked to a DAG backbone containing on position sn-2 the uncommon cis-vaccenic acid . Despite containing a glucose sugar , Glc-DAG-s2 was at least as active as the Borrelia BbGL-2c lipid in activating a mouse iNKT cell hybridoma and it showed similar antigenic potency in vivo ( Figure 4B; [8] ) . Interestingly , we show here and in the previous studies that these unusual features are required together for the glycosylated DAG lipid to have any measurable antigenic potency . These stringent requirements were correlated with unusual binding kinetics compared to the B . burgdorferi DAG antigens , characterized by slow association and slow dissociation rates of the mCD1d-ligand complex to the iNKT TCR [8] . The structure of the mCD1d-Glc-DAG-s2-TCR complex presented here , together with the other studies we have carried out , allow us to understand the stringent chemical requirements , as well as the distinct TCR binding kinetics , in the recognition of the DAG antigens from these highly pathogenic bacteria . As for the other DAG antigen , BbGL-2c , the iNKT TCR is able to induce conformational changes on both the Glc-DAG-s2 ligand and mCD1d ( Figures 2 and 5A ) , which result in a conserved binding mode , as well as a weaker affinity , typical of the DAG antigens compared to their glycosphingolipid counterparts [14] , [22] . However , due to the presence of glucose , with a different conformation of its 4′ hydroxyl group compared to galactose , a hydrogen bond with the CDR1α of the TCR is lost , while a new contact with Gly155 on the α2 helix of CD1d is formed . While this alteration does not translate to an overall change in TCR affinity at equilibrium , it has profound effects on the kinetics of binding of the antigen complex with mCD1d by the iNKT cell TCR . Previous studies showed that contacts between the ligand and the TCR dominate the initial association phase , while protein-protein ( and ligand-protein ) contacts affect the stability of the complex [9] . Therefore , the loss of a contact with the α chain of the TCR can account for the slower TCR association rate exhibited by this ligand . Interestingly , the iNKT TCR appears to be especially sensitive to the conformation of the 4′-OH group , with glucose-containing antigens showing generally reduced potency ( in terms of cytokine release by iNKT cells ) [9] , [10] and preferential proliferation of Vβ7+ cells [10] compared to α-GalCer . Furthermore , the consequent locking of the glucose head in the favorable position following TCR engagement , described here for Glc-DAG-s2 , likely contributes to a slower dissociation . These novel contacts rely on the presence of both the vaccenic acid and glucose , as the variants with an oleic acid in place of the vaccenic acid , or a galactose replacing the glucose , are considerably less active . Consistent with this , a model of Gal-DAG-s2 suggests that the presence of an axial 4′ hydroxyl would be located in an unfavorable position for recognition by the iNKT TCR ( Figure S2 ) . Moreover , the antigenicity of the ligand requires vaccenic acid to be in the sn-2 position of the ligand in order to orient correctly the glucose for recognition by the TCR . Interestingly , these structural requirements do not appear to be influenced by the variable CDR3β loop , as three different hybridomas responded to Glc-DAG-s2 at comparable levels ( Figure 5C ) . Glc-DAG-s2 also stimulates human iNKT cells [8] but is not clear whether the same stringent requirements observed in mouse are conserved in the human CD1d-TCR interaction as no structural information is available on the modality of recognition of DAG antigens by the human iNKT TCR . Clearly , more work has to be done to illuminate the structural basis of microbial DAG recognition by human iNKT cells . Consistent with a model in which the TCR contacts first the ligand and subsequently CD1d , our mutational data also show that the protein-protein interface above the F′ pocket is critical for the interaction , and specifically , that this region determines the dissociation rate , and therefore the stability , of the mCD1d-TCR complex ( Figure 6 ) . Interestingly , the mechanism of antigen recognition by the iNKT TCR appears to be radically different to what has been observed for MHC-TCR interactions , where the TCR first contacts residues on the antigen presenting molecule and only a later stage contacts the antigen itself [23] . The extensive amount of structural and biochemical information recently collected on the interaction between CD1d and the iNKT TCR is consistent with the idea of the iNKT TCR as a pattern recognition receptor [10] , [12]–[14] , [18] . While the S . pneumoniae antigen follows the typical pattern of an α-linked sugar to a diacyl backbone , the data presented here show clearly that , within this pattern , stringent requirements are still in place . In particular , the Glc-DAG-s2 ligand exemplifies the case of a relatively weak hexose and an uncommon alkyl chain synergistically contributing to the potency of an iNKT antigen .
The expression and purification methods of fully glycosylated mouse CD1d/β2m heterodimer proteins were reported previously [11] . Mouse TCR refolding was performed according to previously reported protocols [14] with minor modifications . 64 mg of α chain and 96 mg of β chain inclusion bodies were mixed together and added drop wise to 1 L refolding buffer ( 50 mM Tris-HCl , 0 . 4 M L-arginine , 5 M urea , 2 mM EDTA , 5 mM reduced glutathione , 0 . 5 mM oxidized glutathione , 0 . 2 mM PMSF , pH 8 . 0 at RT ) two times . The refolding mix was dialyzed twice against 18 L dialysis buffer 1 ( 10 mM Tris-HCL , 0 . 1 M urea , pH 8 . 0 ) for 16 h and then once against 18 L of 10 mM Tris-HCl pH 8 . 0 for 24 h . The refolded TCR proteins were purified by MonoQ 5/50 GL ( GE Healthcare ) using a linear NaCl gradient ( 0–300 mM NaCl ) followed by size exclusion chromatography using a Superdex S200 10/300 GL ( GE Healthcare ) in 50 mM Hepes pH 7 . 5 , 150 mM NaCl . The synthetic DAG ligand Glc-DAG-s2 was synthesized as previously reported [8] and dissolved at 4 mg/ml in DMSO . mCD1d was incubated overnight with 3–6 molar excess of Glc-DAG-s2 in presence of 0 . 05% Tween-20 and 100 mM Tris-Cl pH 7 . 0 . Glc-DAG-s2 loaded CD1d was purified by size exclusion chromatography first and then incubated with equimolar amount of TCR for 30 min without further purification . The complex was concentrated to 4 . 8 mg/ml for crystallization . Crystals of mCD1d-Glc-DAG-s2-TCR complexes were grown at 22 . 3°C by sitting drop vapor diffusion while mixing 0 . 5 µl protein with 0 . 5 µl precipitate ( 17% polyethylene glycol 3350 , 8% v/v Tacsimate pH 5 . 0 ) . Crystals were flash-cooled at 100 K in mother liquor containing 20% glycerol . Diffraction data were collected at the Stanford Synchrotron Radiation Laboratory ( SSRL ) beamline 7 . 1 and processed with the iMosflm software [24] . The mCD1d-Glc-DAG-s2-TCR crystal belongs to space group C2221 with cell parameters a = 78 . 1 Å; b = 190 . 7 Å; c = 150 . 9 Å . The asymmetric unit contains one mCD1d-glycolipid-TCR molecule with an estimated solvent content of 55 . 0% . The structures were determined by molecular replacement using MOLREP as part of the CCP4 suite [25] , [26] using the protein coordinates from the mCD1d-iGb3 structure ( PDB code 2Q7Y ) [27] , followed by the Vα14Vβ8 . 2 TCR [14] ( from PDB code 3O8X ) as the search model . When a MR solution containing both mCD1d and TCR was obtained , the model was rebuilt into σA-weighted 2Fo–Fc and Fo–Fc difference electron density maps using the program COOT [28] . Maximum-likelihood restrained refinement coupled with TLS refinement was performed in REFMAC [29] with five anisotropic domains ( α1-α2 domain of CD1d , including carbohydrates and glycolipid , α3-domain , β2m , variable domains and constant domains of the TCR ) . The quality of the model was evaluated with the program Molprobity [30] and the validation tools available in COOT . Shake-omit maps were generated by removing the ligand from the structure and randomly perturbating the coordinates , occupancy , and B-factor of each atom by 0 . 2 Å , 0 . 05 units , and 20 Å2 , respectively , with the software Moleman2 [31] . The resulting structure was then refined with the software REFMAC as described earlier . Data collection and refinement statistics are presented in Table 1 . Coordinates and structure factors have been deposited in the Protein Data Bank under accession code 3TA3 . Mouse CD1d mutants were generated using Quick Change II Site-Directed Mutagenesis Kit ( Stratagene , Agilent Technologies ) according to the manufacturer's instructions with the primers indicated below . Mutated constructs were purified with the Qiagen Miniprep Kit ( Qiagen ) and the presence of the mutation confirmed by sequencing . The mutated birA-tag mCD1d/β2m were expressed and purified using the same method described above for mCD1d/β2m . Primer sequences: L84V 5′-ttaccagggacatacaggaagtagtcaaaatgatgtcacc-3′; L84V_antisense 3′-aatggtccctgtatgtccttcatcagttttactacagtgg-5′; L84F 5′-accagggacatacaggaattcgtcaaaatgatgtcacc-3′; L84F_antisense 3′-tggtccctgtatgtccttaagcagttttactacagtgg-5′; V149L 5′-cttggttagacttgcccatcaaattgctcaacgctg-3′; V149L_antisense 3′-gaaccaatctgaacgggtagtttaacgagttgcgac-5′; L150V 5′-cttgcccatcaaagtggtcaacgctgatcaagg-3′; L150V_antisense 3′-gaacgggtagtttcaccagttgcgactagttcc-5′; M69A 5′-gtgggagaagttgcagcatgcgtttcaagtctatcgagtc-3′; M69A_antisense 3′-gtgggagaagttgcagcatgcgtttcaagtctatcgagtc-5′; M162A 5′-caagtgcaaccgtgcaggcgctcctgaatgacacct-3′; M162A_antisense 3′-caagtgcaaccgtgcaggcgctcctgaatgacacct-5′ . A20/CD1d cells are derived from murine B cell lymphoma A20 ( American Type Culture Collection , Rockville , MD ) , with stable expression of wild type mouse CD1d [7] , [32] . A20/CD1d and iNKT hybridomas cell lines Hy2C12 , Hy1 . 2 , and Hy1 . 4 were cultured in RPMI 1640 medium supplemented with 2 mM L-glutamine , 100 mg/ml each of penicillin and streptomycin , 50 mM 2-mercaptoethanol , and 10% FBS . Mouse iNKT cell hybridoma 1 . 2 ( Vα14/Vβ8 . 2 ) has been described previously [3] , [6] . 1×106 A20/CD1d cells expressing wild-type mCD1d were cultured in complete medium containing indicated amounts of lipid antigens or vehicle ( 56 mg/ml sucrose , 7 . 5 mg/ml histidine , and 5 mg/ml Tween-20 [pH 7 . 2] ) overnight . On the second day of culture , A20/CD1d were collected , washed thoroughly , and 1×105 APCs were seeded in the presence of 5×104 iNKT cell hybridomas per well in a 96-well plate for 24 h , and IL-2 in the supernatant was measured by ELISA according to the manufacturer's instructions ( BD Biosciences ) . Stimulation of mouse iNKT cell hybridomas on microwell plates coated with soluble mCD1d was carried out according to published protocols [3] , [6] , [33] , with a few modifications . Briefly , the indicated amounts of compounds were incubated for 24 h in microwells that had been coated with 1 . 0 µg of mCD1d . After washing , 5×104 iNKT cell hybridoma cells were cultured on the plate for 16 h , and IL-2 in the supernatant was measured by ELISA according to the manufacturer's instructions ( R&D systems ) . Surface Plasmon Resonance studies using a refolded and biotinylated Vα14Vβ8 . 2 TCR were carried out as previously reported [11] with 300–500 response units ( RU ) of biotinylated mCD1d-vehicle or mCD1d-ligand immobilized on the chip . Serial dilutions of Vα14Vβ8 . 2 were injected with increasing concentrations ( 0 . 002–1 . 25 µM ) over a streptavidin CAPture chip ( GE Healthcare ) . The experiment was performed twice . Loading efficiency was measured by immobilizing the biotinylated mCD1d-ligand ( after incubation for 16 h in the presence of 1 µg/ml of α-GalCer ) complex on a CAPture chip ( 400–500 RU ) followed by the injection of a saturating concentration ( 1 µM ) of the Fab portion of the mCD1d-α-GalCer specific antibody L363 [20] . 100% glycolipid loading efficiency is achieved when the increase in RU upon Fab binding is equal to the RU of CD1d-glycolipid coated on the chip , as mCD1d-glycolipid and Fab have a comparable molecular weight . | Invariant natural killer T ( iNKT ) cells are an evolutionarily conserved population of immune cells that recognize lipid antigens . A protein called a T cell receptor for antigen ( TCR ) on the surface of these iNKT cells recognizes lipids bound to a protein called CD1d on the surface of antigen-presenting cells . Here we describe the three-dimensional structure of the complex that forms between CD1d and the iNKT TCR together with a glycolipid antigen from the infectious bacterium Streptococcus pneumoniae , which is a common cause of bacterial meningitis in adults and is responsible for many other pneumococcal infections . We determined the three-dimensional structure of the complex by X-ray crystallography . The data obtained allow us to understand the structural requirements that make this glycolipid a potent antigen for iNKT cells , and why the TCR of these cells recognizes a particular combination of hexose sugar and diacylglycerol lipid . Moreover , by mutating CD1d and using biophysical methods to study the mutant protein complexes , we analyzed the role of the protein–protein interface between CD1d and the TCR and found that it plays an important role in the stability , but not the formation , of the trimolecular complex containing glycolipid antigen . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"immunology",
"biology"
] | 2011 | Unique Interplay between Sugar and Lipid in Determining the Antigenic Potency of Bacterial Antigens for NKT Cells |
The effect of drugs , disease and other perturbations on mRNA levels are studied using gene expression microarrays or RNA-seq , with the goal of understanding molecular effects arising from the perturbation . Previous comparisons of reproducibility across laboratories have been limited in scale and focused on a single model . The use of model systems , such as cultured primary cells or cancer cell lines , assumes that mechanistic insights derived from the models would have been observed via in vivo studies . We examined the concordance of compound-induced transcriptional changes using data from several sources: rat liver and rat primary hepatocytes ( RPH ) from Drug Matrix ( DM ) and open TG-GATEs ( TG ) , human primary hepatocytes ( HPH ) from TG , and mouse liver / HepG2 results from the Gene Expression Omnibus ( GEO ) repository . Gene expression changes for treatments were normalized to controls and analyzed with three methods: 1 ) gene level for 9071 high expression genes in rat liver , 2 ) gene set analysis ( GSA ) using canonical pathways and gene ontology sets , 3 ) weighted gene co-expression network analysis ( WGCNA ) . Co-expression networks performed better than genes or GSA when comparing treatment effects within rat liver and rat vs . mouse liver . Genes and modules performed similarly at Connectivity Map-style analyses , where success at identifying similar treatments among a collection of reference profiles is the goal . Comparisons between rat liver and RPH , and those between RPH , HPH and HepG2 cells reveal lower concordance for all methods . We observe that the baseline state of untreated cultured cells relative to untreated rat liver shows striking similarity with toxicant-exposed cells in vivo , indicating that gross systems level perturbation in the underlying networks in culture may contribute to the low concordance .
Transcriptional changes in model systems are often used to elucidate mechanism-based effects of drug treatment and the relevance for humans [1 , 2] . While the use of model systems is viewed with skepticism by some , it is common practice to use nonclinical species to inform our understanding of human response ( e . g . , mouse knockout to human ) , to extrapolate effects in cell lines to more complex tissues , ( e . g . , transformed cell lines to tumors ) , or to use observations in cultured primary cells to understand the behavior of cells in situ in the organ of interest for a disease process ( e . g . , primary hepatocytes to liver ) . In each case it is assumed that effects in a less complex system are relevant to a more complex system . Further complicating the use of gene expression to solve these multi-scale and multi-dimensional problems are technical challenges imposed by variability across measurement platforms [3 , 4] and differences in experimental protocols ( e . g . dose and time ) which hamper the ability to aggregate data from multiple sources thereby reducing applicability of existing data . Methods that measure the effects of perturbations across multiple genes , such as gene set analysis ( GSA ) [5] and co-expression networks [6] may reduce technical and experimental noise while boosting relevant biological signals across data sources [7] . Gene expression profiling has been applied in many areas . Concerns over drug safety , and in particular drug-induced liver injury , have resulted in large projects to establish reference gene expression databases in nonclinical species [8 , 9] . Prediction of liver toxicity in humans from nonclinical species is of particular interest and the rat is a commonly used nonclinical species for testing safety prior to clinical development . Thus , measured by number and diversity of drugs , the rat liver is by far the most extensively studied in vivo model using gene expression profiling . Calls to eliminate animal tests in favor of human in vitro models increase the need to understand the relevance of conclusions from gene expression studies across models and species [10–12] . Several small scale studies have compared in vitro vs . in vivo gene expression profiles following drug treatment [13–17] . Discrepancies observed in those studies have been attributed to pharmacokinetics [18 , 19] and differences in the baseline state of liver vs . its constituent cells in culture [20–22] . However , there are no large scale studies that compare the concordance of in vitro and in vivo response of hepatocytes to drug exposure . Previous studies have examined the concordance of gene-expression profiles generated using the same samples analyzed with different microarrays and/or RNA-seq [23 , 24] . Since the regulation of gene expression varies across tissues and organisms ( “models” ) [25] , it is also of interest to compare concordance across models . In this work , we take a systems level approach to examine the concordance of transcriptional effects for similar treatments taken from different sources for the same model ( i . e . different laboratories ) , across species ( rat , mouse , human ) and across cell models comparing in vitro to in vivo . Leveraging the wealth of time-resolved data within the TG-GATEs database [9] , the effect of drug treatment on rat liver is compared to effects observed in mouse liver , cultured rat primary hepatocytes ( RPH ) , cultured human primary hepatocytes ( HPH ) , and HepG2 cells . We compare results using gene expression profiles for individual genes , gene sets or co-expression networks ( modules ) . Modules have been widely used for disease characterization studies [26–29] and reduce the dimensionality of gene expression data while avoiding biases inherent in relying on canonical pathways that dominate gene set enrichment analysis [5] . We examine the absolute level of concordance achieved for each method to understand the impact on mechanistic interpretation ( “the most highly induced genes/pathways/modules in RPH are A , B , C and the same genes/pathways/modules are also highly induced in rat liver” ) , and relative concordance ( “are the profiles for a drug in RPH and liver more similar than profiles involving different drugs ? ” ) . We refer to the latter as success in “self-identification”; in the absence of high absolute concordance one may still successfully infer properties of a drug ( classification ) via those of drugs having the most similar expression profiles ( i . e . ConnectivityMap-type applications [30–33] ) . We explore possible causes for the low concordance of in vivo vs . in vitro profiles for all methods by comparing the baseline state of cultured cells vs . rat and human liver as an estimate of system perturbation . By analyzing whole liver vs . freshly isolated and de-differentiating RPH after various times in culture , we dissect the contributions of non-parenchymal cells resident in liver tissue vs . changes associated with de-differentiation in culture as factors explaining divergence from rat liver gene expression .
An extensive characterization of transcriptional effects of drugs and other compounds in rat liver and related model systems have been compiled in two publicly-available resource: the Drug Matrix ( DM ) [8 , 34] and the open TG-GATEs ( TG ) databases [9] . We analyzed Affymetrix microarray data for 3528 experiments in male rat liver from TG , and 654 similar experiments from DM . An experiment denotes expression results for a unique combination of species , strain , tissue , drug , dose and time , usually obtained by comparing 3 biological replicates in the treatment group to 3 or more replicates in the control group . Treatment groups are always time-matched to controls . To establish concordance with rat liver results across different models , we analyzed 1260 experiments performed in RPH and 941 experiments in HPH from TG , and 268 RPH experiments from DM . We also retrieved 49 mouse liver and 177 HepG2 experiments from the Gene Expression Omnibus ( GEO ) repository [35] , most of which overlap with drugs available in TG and DM . The experiments considered for this work are reported in S1 Dataset . Many approaches have been employed for describing drug-induced transcriptional changes at various levels of granularity , including gene-level , gene set analysis ( GSA ) and co-expression network analysis . Transcript abundance is a well-established factor governing the concordance of microarray and RNA-seq analyses [23] . To avoid including poorly-performing low abundance transcripts , we identified 9071 genes , henceforth described as “liver-expressed genes” having above median intensity in ≥10% of DM rat liver treatment or control arrays . Fold-change values were calculated for each gene by comparing treatment and control samples . In this work , the ensemble of 9071 fold change values is denoted as gene-level analysis . In addition , we performed GSA ( using the PAGE algorithm [36] ) to identify gene sets enriched among perturbed genes for each experiment , using a subset of the MSigDB collection [5] . Finally , we calculated the eigengene score ( or module score ) for each of 415 co-expression networks obtained with the WGCNA package [37] applied to the DM rat liver dataset . The eigengene summarizes induction / repression of the underlying module genes , with gene weights proportional to their degree of co-expression . Individual genes , gene sets and modules are described as “features” below . To evaluate the concordance of transcriptional changes caused by drug treatment , assessed via gene-level , gene-set and module analysis , we first identified all pairs of experiments that involve the same drug . Each experiment in the pair may involve a different dose , treatment time , system or source of data . Two metrics commonly used in comparing gene expression results were used to assess concordance: the Pearson correlation coefficient and percent overlap among the top 5% of differentially-expressed features ( genes , gene sets or modules ) . Percent overlap was employed for this purpose in the MicroArray Quality Control projects [23 , 24] . Fig 1 illustrates the comparison of three methods and two metrics on azathioprine treatment in rat liver and RPH . Comparison of the metrics in the TG data shows that Pearson R and the 5% overlap metric are only approximately correlated . The percent overlap at other thresholds ( top 2 . 5% and top 10% ) yield similar results to the 5% threshold ( S1 Fig ) , henceforth , the overlap metric for all analyses uses the top 5% of features . Within the in vivo or in vitro setting , there are few drugs studied at the same dose and time in multiple models or sources ( 13 for TG vs . DM rat liver , 2 for TG vs DM RPH ) . This severely limits our ability to study concordance across models and sources for identical treatments . We sought to identify factors with small effects on concordance within one system and source , in order to relax our criteria for identifying comparable experiments across models and sources . To this end , we tabulated the concordance of 78 , 226 pairs of rat liver experiments and 10 , 008 pairs of RPH experiments from TG involving the same drug . The importance of dose and time difference , study design ( repeat vs . single-dose ) and overall transcriptional activity caused by treatment were studied using stepwise linear regression ( i . e . variables which explain more variation in concordance are those with greater effect on gene expression changes resulting from treatment ) . Across rat liver and RPH , for any of the three analysis methods or two concordance metrics , differences in time and overall transcriptional activity of the least perturbing treatment explain the largest percentage of variation in concordance ( Table 1 ) . Dose differences between treatments explain less than 0 . 1% of variation in all cases . This suggests that rigorous comparisons across sources ( within the in vivo or in vitro setting ) should minimize differences in time , but allow mismatch on dose in order to increase the number of comparable experiments . Further , the level of concordance was strongly influenced by the level of transcriptional activity for the least perturbing of two treatments , as quantified by the average absolute eigengene score across the co-expression modules . Thus , concordance for subtle treatment-induced transcriptional effects is lower than for treatments causing large effects . Our selected measure of transcriptional activity , the average absolute eigengene score ( avg . abs . EG ) , is highly correlated with percent of genes differentially expressed ( S2 Fig ) . Similar conclusions are reached regarding the impact of dose and time differences by tabulation of concordance for adjacent doses or time points in the TG-GATEs study design ( S1 Table ) . Having identified dose differences as being of lesser importance in determining concordance within the TG rat liver system , we identified comparable experiments in order to assess concordance across sources and systems . The TG rat liver model serves as reference system for comparing drug-induced transcriptional responses in other model systems . We define three levels of stringency in identifying comparable treatments in liver: 1 ) no time difference and ≤ 5-fold dose difference; 2 ) ≤ 2-fold time difference and ≤ 10-fold dose difference; 3 ) no constraint on time or dose differences . Within each comparison category , we identify the most concordant TG rat liver experiment satisfying the constraints . This selection is repeated for both concordance metrics and 3 analysis methods ( genes / GSA / modules ) . Level 2 is arguably a reasonable choice for assessing cross-source concordance , as pharmacokinetic parameters vary across individuals and species . Level 3 analysis offers a very optimistic view of concordance , whereby each experiment is compared to all TG rat liver experiments for the same drug and the most concordant pair retained . Since all drugs were tested at 3 doses and 3 , 6 , 9 and 24 hours , and most drugs were also tested at 4 , 8 , 15 and 29 days , this view takes the highest level of concordance across 12 or 24 in vivo “snapshots” from TG liver for a given drug . By necessity , we allowed ≤ 1 . 15-fold time differences ( 7 vs . 8 day study comparisons , required for GSE43977 ) for level 1 comparisons from rat to mouse liver , because few experiment pairs across models used exactly the same time point . Within the constraints defined above , we performed three comparisons using liver transcript profiles: 1 ) TG rat liver vs . itself , 2 ) TG rat liver vs . DM rat liver and 3 ) TG rat liver vs . GEO mouse liver . The first comparison serves as a positive control . This establishes the upper boundary of concordance that can be expected within a large dataset built using a uniform protocol but with information collected over an extended period of time or using different staff and facilities , factors know to influence concordance across microarray experiments even when the same samples are used [3 , 4] . The second comparison serves as a robust assessment of cross-source concordance across 33 shared drugs having experiments in both DM and TG liver that match at the highest level of stringency . This represents a “real-world” assessment of concordance , as the DM and TG efforts used the same rat strain ( Sprague Dawley ) but sourced from different colonies ( Charles River Japan vs . Wilmington , MA ) with slightly different ages at study onset ( 6 weeks vs . 7–9 weeks ) , housed/handled in different facilities and non-identical microarray study protocols [8 , 9 , 38] . The third comparison begins to address the extent to which gene expression changes observed in one species are relevant to another , by examining transcriptional effects of 10 drugs for which we could find mouse liver experiments matching at the high stringency level in GEO . In addition to analysis of transcriptional effects via genes , GSA and modules , we considered GSA applied exclusively to REACTOME pathways , and co-expression network analysis using a subset of 216 modules with preservation Z-summary ≥ 5 in TG rat liver ( methods ) . The latter represent the ~50% of modules derived in DM rat liver with the highest degree of validation in an independent dataset . As above , concordance between pairs of experiments is evaluated using the Pearson R and overlap metric . In addition , we report the probability that a given level of concordance between experiments involving the same drug exceeds the level seen for random pairs involving different drugs ( methods ) . This is especially relevant for CMap-type analyses , where the goal is not mechanistic understanding of drug effect , but the inference of drug properties via those of well characterized drugs with similar gene expression profiles [30–33] . For simplicity below , we refer to this classification metric as successful “self-identification” vs . random pairs involving different drugs: the gene expression profile for a given drug is more similar to other profiles involving the same drug ( in another system , or at a different dose or time ) than profiles involving different drugs . Concordance of experiment pairs is tabulated across quartiles of transcriptional activity for the least-perturbing of two treatments in the pair , given the importance of this variable in explaining intra-source concordance ( Fig 2 , S2 Dataset ) . This can be seen from comparisons within TG rat liver , where all methods yield higher concordance with increasing levels of transcriptional activity ( Fig 2A ) . Improvement in concordance when relaxing constraints on dose and time differences ( pink -> blue -> green bars ) is minimal within-source , indicating that the most similar experiments by gene expression tend to be the most similar experiments by dose/time . All methods have high self-identification probability . The TG vs . DM liver comparison reveals lower concordance than the within-source trends described above , coupled to greater improvement in concordance when relaxing constraints on dose and time differences between experiments ( Fig 2B ) . The dependence on level of transcriptional effects is larger , with ca . 50% or less success at self-identification in the lower quartiles . The comparison between TG rat and GEO mouse liver is broadly similar for GSA and modules , especially when allowing less stringent matches on dose and time ( Fig 2C; pink+blue bars ) . Gene-level concordance is lower relative to GSA or modules for the cross-species comparison on both metrics ( within each of 18 combination of two metrics , 3 ranges of avg . abs . EG and 3 dose/time stringency levels , p < 0 . 05 for 9 GSA:all vs . gene comparisons and 12 module:all vs . gene comparisons using one-sided t-tests with unequal variance ) . In contrast , the high performance of gene-level analysis for self-identification between rat and mouse liver indicates that appropriate inferences are made by CMap analyses ( i . e . low absolute concordance , but high enough to associate the same or similar drugs across models ) . The importance of considering both the quantitative level of concordance , along with probability of self-identification is illustrated in S3 Fig . The threshold of similarity required to exceed that observed for random experiment pairs differs for each combination of method and metric . This threshold also increases as a function of overall transcriptional activity , since the concordance for random pairs involving different drugs increases as a function of avg . abs . EG ( S2 Table ) . For experiments in rat liver in the mid-range of transcriptional activity , the level of similarity on the Pearson metric required to achieve a threshold of 50% success at self-identification increases from approximately 0 . 2 for gene-level analysis to 0 . 55 for GSA:REACTOME . As such , the apparent advantage of GSA:REACTOME on the percent overlap metric ( Fig 2 ) is not sustained when considering success at self-identification ( S2 Dataset ) . A significant challenge in comparing results from gene expression profiling between in vivo and in vitro experiments arises from the confounding effects of in vivo pharmacokinetics . This is elegantly described for methapyrilene-treated rat liver and rat hepatocytes by Schug et al . , where concentration decreases slowly in culture but rapidly ( and at different rates ) in blood from arterial , liver and portal veins [19] . A further dilemma in aligning in vivo vs . in vitro exposures concerns differences in non-specific binding between plasma and media [39] , neither of which have been measured in the TG-GATEs initiative . We adopted the use of an optimistic comparison , whereby each in vitro experiment is compared to all TG rat liver experiments for the same compound and the most concordant pair retained . We assume that one or more of these 12 or 24 in vivo “snapshots” should reasonably reflect the drug / time conditions under study in vitro . The approach is illustrated for 4 μM azathioprine treatment in RPH 24 hours after drug administration . When using the Pearson metric , the most similar in vivo profile is the high dose at 6 hours for gene and module analysis , vs . the 9 hour time point when using GSA ( Fig 3 ) . Using the overlap metric selects a similar in vivo condition for GSA and modules , but the 8 day condition for gene-level analysis . We selected azathioprine as a case where the correlation ( using the Pearson metric ) oscillates from positive to negative to minimal , a behavior that reflects the dynamic nature of biological response patterns in vivo for acute vs . chronic dosing . This example serves to illustrate a key difference between the metrics: Pearson correlation can be negative , conveying a reversal of states ( which has been successfully used for drug repurposing[31] ) , while percent overlap is limited to a range between 0 and 100 and does not distinguish a situation where no genes overlap between two experiments from one where they overlap strongly but change in different directions . We used the TG rat liver and RPH experiments to assess whether drug response in cultured rat hepatocytes and rat liver are similar . This comparison is of high interest since the drug lots and experimental procedures for microarray analysis are the same in both experimental models , minimizing the role of technical differences in our assessment . For each of 1255 TG RPH experiments involving 140 unique drugs , we determined the most similar TG rat liver experiment and tabulated the average concordance across methods ( Fig 4A ) . We analyze separately a subset of 207 modules with preservation score ≥ 3 in TG RPH , representing ~50% of modules with higher confidence of co-regulation in culture . Concordance measured on both metrics is lower than the equivalent within-source comparisons for TG rat liver ( Fig 2A ) and TG RPH ( S4A Fig ) . As observed for rat vs . mouse liver , GSA and modules outperform gene-level analysis; differences for both metrics in each of the 3 ranges of avg . abs . EG are all statistically significant at p < 1e-11 when comparing GSA and modules to genes ( one-sided t-tests with unequal variance ) . Low in vivo vs . in vitro concordance is also seen when comparing 114 DM RPH experiments to TG liver ( S4C Fig ) . We investigated whether concordance varied across pharmacological classes represented with 3 or more drugs in TG using modules preserved in RPH , the top performing method when considering absolute concordance and probability of success at self-identification ( S3 Table ) . Although we find little evidence of concordance variation across classes , this observation should be tempered by the very limited redundancy of drug classes in TG ( and the diversity within classes , e . g . grouping together all non-steroidal anti-inflammatory drugs ) . The highest in vivo vs . in vitro similarity is achieved for cycloheximide , galactosamine , tunicamycin , ethionine , azathioprine and 1-napthtyl isocyanate , a group of structurally and pharmacologically diverse agents . Lower concordance between rat liver and RPH could arise from a combination of pharmacokinetics or differences in the baseline states of hepatocytes affecting how they respond to perturbation . Comparisons involving different in vitro models mostly eliminate the impact of pharmacokinetics , as drug exposures from common in vitro experiments do not change substantially over 24 hours ( compound dissolved in culture media serves as a buffer ) [19] . We compared drug-induced expression changes between TG RPH and TG human primary hepatocytes ( HPH ) , comparing each HPH experiment against 12 TG RPH experiments ( 3 doses × 4 time points ) . As done for the liver comparisons , concordance is assessed at 3 levels of restriction on dose and time differences , and the most favorable comparison retained at each level ( Fig 4B ) . A similar analysis was performed by comparing 25 HepG2 experiments from GEO involving 18 drugs to 12 TG HPH experiments for each drug ( Fig 4C ) . Concordance for all methods is low , whether considering absolute concordance or success at self-identification . This result is unlikely to stem from differences in cytotoxicity at a given concentration , since most drugs tested in our hands exhibit similar effects on cell viability in RPH and HepG2 cells ( S4 Table ) . A very limited comparison between rat and mouse primary hepatocytes for 5 drugs , used at similar doses for the same treatment duration , yields qualitatively similar conclusions: the most concordant profiles are those from treatments associated with higher levels of transcriptional activity , yet concordance is generally low across these in vitro models ( S5 Table ) . Given that within-source comparison performed well for all analysis methods , we investigated other sources of variation across experimental models by comparing their baseline states , i . e . basal level of gene expression before perturbation with drug treatment . We find that baseline gene expression in control samples is highly correlated across sources for the same system ( TG vs . DM rat liver , TG vs . DM RPH; Tables 2 and S6 ) . We also find higher correlation between expression in liver and primary hepatocytes of the same organism in culture versus liver-to-culture expression comparisons across organisms: i . e . , rat liver expression is more correlated with RPH expression , than with mouse or human liver; likewise for human liver vs . HPH and mouse liver vs . MPH . However , viewing differences in relative terms ( i . e . , treating liver as ‘control’ and culture as the ‘perturbation’ ) reveals that the transcriptional impact on hepatocytes in going from liver to culture is comparable in magnitude to the effect of administering highly toxic treatments to rat liver causing marked changes in liver morphology ( Fig 5 ) . It is noteworthy that the comparison of GEO mouse liver to TG rat liver appears less dramatic on this basis ( avg . abs . EG = 1 . 75 ) compared to the baseline expression correlation . Thus even in cases where simple gene level correlation suggest that two models are similar , the underlying degree of perturbation reflected by changes in co-regulation behavior of genes , and by analogy the biology associated with those genes , are highly perturbed . To understand the causes and nature of the underlying perturbation reflected in baseline perturbations noted for the in vitro models , we investigated both the contributions of non-parenchymal cells to in vivo transcript profiles and changes associated with the adaptation of the hepatocytes to culture , often characterized as a de-differentiation process [40 , 41] . Non-parenchymal cells comprise approximately 6 percent by volume of the liver ( 40% by cell count ) [42] and cell population differences between liver and enriched freshly isolated hepatocytes may account for some level of discrepancy between the transcriptional response in whole liver vs . response in cultured hepatocytes . We first prepared enriched rat hepatocytes , using a standard liver perfusion protocol , and determined the transcript profiles immediately after isolation ( 0 hour time point ) , and after 4 , 24 and 48 hours of culture on collagen-coated plates . We compared the hepatocyte transcript profiles to those from 3 samples of intact liver serving as a control . We performed two normalizations for the 4 , 24 and 48 hour samples: 1 ) using control liver to assess the impact of a reduction in non-parenchymal cells relative to normal liver; and 2 ) using the 0 hour isolated hepatocyte time point as control to determine change associated with alteration in hepatocellular phenotype in culture . Both module or gene-level analysis demonstrate high concordance between the responses noted in 24 or 48 hour RPH vs . liver in our experiment , and those obtained by comparing vehicle-treated RPH and liver control samples in DM and TG , despite differences of time in culture , culture protocols and the use of vehicle treatment for the DM and TG samples ( Fig 6; S3 Dataset ) . This suggests a high degree of similarity between the underlying biological perturbations in our experiment and those in both DM and TG studies . The transcriptional impact of removing the bulk of non-parenchymal cells ( 0 time vs control liver ) is relatively large in the context of rat liver treatments ( avg . abs . EG 0 . 64 , or 81th percentile among 4182 DM and TG rat liver experiments ) . However , equivalent comparisons at 4 , 24 and 48 hour samples indicate substantially larger perturbation ( avg . abs . EG scores of 2 . 1 , 2 . 4 and 2 . 5 , respectively , ≥99th percentile ) . Differentially expressed genes at the 0 hour are consistent with attribution to cell type differences since a variety of pathways and GO-biological process terms not generally associated with hepatocytes were enriched in down-regulated genes , e . g . hemoglobin complex , extracellular matrix , immune response , leukocyte migration ) . Liver is a site for extramedullary hematopoiesis ( hemoglobin complex ) , stellate cells remodel extracellular matrix components in liver , and there are a variety of liver resident immune cells among the non-parenchymal liver cell population [42] . Pathways and terms enriched at the 4 , 24 and 48 hour time points indicate dramatic differences in activity of fundamental cellular process reflecting cell activity , movement and cytoskeletal structure often associated with de-differentiation ( e . g . up-regulation of RNA processing , ribosome biogenesis , translation , focal adhesion and microtubule cytoskeleton; down-regulation of fatty acid oxidation , CYP450 drug metabolism; S4 Dataset ) . Changes in expression at 24 or 48 hours in RPH culture are highly correlated , whether they are compared to the 0 hour time point or rat liver , suggesting the cells stabilize after 24 hours in culture . Taken together , this indicates that de-differentiation of hepatocytes in cultures accounts for most of the difference relative to rat liver rather than differences in cell type composition . Further , effects observed in comparing cultured rat hepatocytes to rat liver are broadly similar to those seen for cultured human hepatocytes vs . human liver and mouse hepatocytes vs . mouse liver ( Fig 6; S3 Dataset ) . This suggests that the processes responsible for hepatocyte de-differentiation in culture are similar across organisms . To determine if the changes seen upon de-differentiation in culture are relevant to any biological states in rat liver , we compared the changes noted in culture to various liver states after drug treatment in TG and DM . Of note , the transcriptional effect of cell culture on hepatocytes is highly correlated with the effect of 24 hours of bortezomib treatment in rat liver ( Pearson r = 0 . 85; Fig 6 ) and with the treatments included in Fig 5 ( Pearson R of 0 . 67 for cycloheximide , 0 . 77 for carbon tetrachloride , 0 . 66 for methapyrilene , and 0 . 69 for aminosalycilic acid ) . These data suggest that the perturbations seen upon adaptation of hepatocytes to culture resemble biological states adopted by hepatocytes as a response to drug-induced liver injury in vivo . Thus , treating hepatocytes in this type of simple culture models may be more analogous to dosing animals with an injured liver than they are to responses in a naïve animal .
Deconvoluting from phenotypic information to mechanistic understanding is an essential component of drug discovery and requires extrapolation of results from simple to more complex systems and from one complex species to another . In the former , reducing a complex system to a testable model , such as cells in culture , yields mechanistic results that are extrapolated back to a complex tissue or whole organism . In the latter , information from experiments in one species , e . g . rats or mice , are used to infer behavior of another complex system or population , humans . These comparisons are complicated by the multi-scale nature of these systems as well as differences in the dimensions of dose and time . Inaccurate translation results in failure to predict biological responses in human subjects [43] . Systems approaches are suited to understanding individual changes , such as changes in mRNA levels , yet there are few studies that address concordance of mechanistic information across levels of complexity . We addressed this problem two ways , by evaluating concordance across the evolutionary dimension ( mouse , rat and human ) and across scales of complexity ( cells to tissues ) in the same species . We defined components of variation ( dose , time , overall transcriptional activity ) and evaluated concordance with differing level of stringency using commonly-applied methods ( gene-level and GSA ) and co-expression network ( module ) analysis . The latter approach has not been described for toxicogenomics studies , but represents a class of methodologies frequently used in disease characterization studies [26–29] . A major goal of gene expression analysis is to assemble a specific list of genes , pathways or modules that are differentially expressed in response to drug treatment and reflect meaningful biological response . Understanding the degree to which the list , and therefore the mechanistic interpretation , would change as a result of using other measurement technology ( microarrays , RNA-seq ) , sample , species or culture conditions is important . Our findings on concordance build upon comprehensive analyses published with regards to measurement technology [3 , 4 , 23 , 24] and are summarized as follows: Several studies have analyzed TG data , with the goal of identifying genes or pathways that may be used to predict a phenotype ( e . g . carcinogenicity ) . Zhang et al . identified 4 acute-response genes that predict the occurrence of cell death in vitro or liver injury in vivo [45] . El-Hachem et al . analyzed the data using GSEA to identify REACTOME pathways which can be modulated across systems and associate with carcinogens or PPAR alpha activators [46] . There is a large body of literature which has successfully demonstrated prediction of in vivo properties , especially carcinogenicity , from in vitro expression results ( e . g . [47–50] ) . The goal of these studies is to identify a small number of genes or pathways that predict a specific in vivo property , with the possibility of constraining the selection to genes that behave consistently across several systems ( e . g . different culture models , different species , etc . ) . Here we examine a different application of gene expression profiling , of relevance to a scenario where one uses gene expression profiling to understand mechanistic effects of a drug not necessarily tied to a specific phenotype . Effects at the level of the whole genome are considered , not just a subset of high performing genes identified in situations where a researcher has the benefit of large numbers of samples available for model training . As such , the findings reported here do not inform on the applicability of predictors derived in one system to other systems . Our analysis of baseline states ( expression level in untreated samples ) underscores the dramatic transcriptional changes in culture , comparable in magnitude to very toxic rat liver treatments . Comparison of various in vitro models , where the impact of pharmacokinetics is minimal , suggests that the different baseline states of underlying biological systems in these models accounts for differing response upon drug treatment . To our knowledge , there are no published datasets describing drug-induced transcriptional effects in human liver , limiting our ability to assess the utility of advanced culture models for in vitro studies on human cells ( e . g . “organ-on-a-chip” , 3D-printed tissues , iPS cells , etc ) . Nonetheless , based on the available rodent data , the scope for improvement over “flat” cell culture ( culture of cells adhered to collagen-coated plates ) is evident . Three dimensional culture models have shown improvement in the concordance vs . liver of cell viability assessments [51–53] or gene expression changes for nanomaterials [54] . Our results suggest that demonstrating improved concordance across a range of drug treatments in the context of the well-studied rodent liver seems warranted . The correlation of transcriptional changes that occur when rat and human hepatocytes differentiate in culture suggest that improvements in modelling the rodent liver using cultured rodent cells may be transferrable to the study of human liver effects using human cells .
The studies on rat primary hepatocyte de-differentiation were conducted in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the U . S . National Institutes of Health , and were approved by the Lilly Animal Care and Use Committee . Animals were given an ip injection of Na pentobarbital ( 275 mg/kg ) and sacrificed via exsanguination . Rodent and de-identified human gene expression data from Drug Matrix , open TG-GATEs and the Gene Expression Omnibus ( GEO ) repository are freely available to the public . No institutional review board approval was sought to analyze those data . Rat hepatocytes were isolated from three male Sprague Dawley ( SD ) rats using the method of Berry and Friend [55] and Seglen [56] . During the isolation and prior to perfusion , a lobe of liver was tied off and homogenized in Trizol for RNA isolation according to the manufacturer’s instructions ( Invitrogen ) . This served as the liver in situ reference sample . Cells were isolated and samples from the cell pellet ( time zero ) and cells cultured for 4 , 24 and 48 hours were placed in Trizol for RNA isolation . Cells were cultured in William’s E media supplemented with glutamate , gentamicin , insulin , transferrin , dexamethasone and serum ( 10% FBS ) for 4 hours and then exchanged with serum free media with the same constituents . Three biological replicates were generated for each group ( one from each rat ) . Each biological replicate was analyzed via 3 technical replicates , for a total of 9 array hybridizations per group . All RNA was cleaned using Qiagen RNeasy columns ( Qiagen ) and evaluated on an Agilent Bioanalyzer ( Agilent ) . Samples for mRNA profiling studies were hybridized to Affymetrix Rat Gene Expression 230–2 arrays according to the standard Affymetrix protocol . Briefly , total RNA was used for preparation of biotin-labeled cRNA . Labeled cRNA was fragmented and used for array hybridization . Arrays were washed and stained with streptavidin-conjugated phycoerythrin on an Affymetrix FS450 Fluidics station . The arrays were scanned on an Affymetrix GeneChip Scanner . A summary of the image signal data , detection calls , and gene annotations for every gene interrogated on the array was generated using Affymetrix GeneChip Command Console MAS 5 . 0 algorithm with all arrays scaled to 500 . One array from each of the 4 and 48 hour time points did not meet standard microarray QC metrics and were excluded from analysis . RNA was then labeled and hybridized to RG230-2 microarrays ( Affymetrix ) according to the manufacturer’s instructions . The data has been deposited in NCBI's Gene Expression Omnibus [35] and are accessible through GEO Series accession number GSE74903 http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE74903 ) . Affymetrix RG230-2 , MG430-2 and HGU133-2 microarray data was obtained from multiple sources . Throughout this work , an “experiment” denotes one species / tissue / drug / dose / time combination , usually analyzed with 3 biological replicates and always compared to a time-matched vehicle or DMSO control . The National Institute for Environmental Health Sciences ( NIEHS ) hosts the Drug Matrix database [8] ( DM ) on its website ( https://ntp . niehs . nih . gov/drugmatrix/index . html ) . We retrieved CEL files for 654 male rat liver and 268 rat primary hepatocyte ( RPH ) experiments on RG230-2 microarrays . The TG-GATEs database ( TG ) consists of approximately 160 drugs and reference compounds used to generate gene expression profiles in rat liver , rat kidney , rat primary hepatocytes ( RPH ) and human primary hepatocytes ( HPH ) [9] ( http://toxico . nibiohn . go . jp/english/ ) . We retrieved CEL files for 3528 male rat liver experiments ( single and repeat dose ) and 1260 RPH experiments on RG230-2 microarrays , and 941 human primary hepatocyte ( HPH ) in vitro experiments on HGU133-2 microarrays . MG430-2 CEL files for mouse liver experiments and mouse primary hepatocytes ( MPH ) , and HGU133-2 CEL files for human liver and HepG2 microarray experiments were retrieved from the Gene Expression Omnibus [35] ( GEO; http://www . ncbi . nlm . nih . gov/geo/ ) or Array Express [57] ( https://www . ebi . ac . uk/arrayexpress/ ) : GSE44783 ( 4 , 15 day repeat-dose studies in male CD-1 mice for 14 drugs ) , GSE43977 ( 7 day repeat-dose studies in male C57BL/6J mice for 17 compounds ) , GSE28878 , GSE51952 ( 72 drugs in HepG2 for 12 , 24 or 48 hours exposure ) , GSE37031 , GSE63067 ( normal human liver ) , GSE57129 , E-MEXP-2539 , E-MEXP-2209 , E-MEXP-2636 ( 19 drugs in MPH for 1 or 2 day exposure ) . Bioconductor version 2 . 13 was used throughout this work ( http://www . bioconductor . org ) . CEL files were analyzed with the Affy package to produce MAS5 and RMA [58] probe set intensities . Normalization to controls was performed by taking the average MAS5 log intensity or average RMA log intensity across 2–5 biological replicates for the treatment group and subtracting the equivalent quantity for the control group . In order to simplify comparisons across system and microarrays , we selected the highest intensity probe set per gene using expression levels in control samples for each combination of system and microarray . The Affymetrix-supplied annotations we used to map probe sets to Entrez gene IDs , retaining only probe sets that map to a single gene ( version 33 for rat liver , RPH and HPH; version 35 for HepG2 and mouse liver . Downloaded from http://www . affymetrix . com/support/mas/index . affx ? navMode=cat120004&aId=supportNav ) . The following data sources were used to select the highest intensity probe set for each gene from MAS5 normalized signals: Drug Matrix control arrays for rat liver and RPH , TG control arrays for HPH , GSE43977 , GSE51885 and GSE44793 male controls for mouse liver ( using both time points and 3 vehicles for GSE44793 , and computing the average intensity across the three datasets giving each equal weight ) , E-MEXP-2209 for MPH ( using all 4 time points for DMSO vehicle ) , GSE28878 for HepG2 ( using all 3 time points for DMSO vehicle ) , GSE37031 and GSE63067 for normal human liver ( excluded GSM1539891 from GSE63067 as it was an outlier when samples were clustered ) . We examined control intensity correlations of the individual GEO series before averaging and all exceed R-sq values of 0 . 88 . While other strategies exist for selecting a representative probe set for each gene , intensity based selection leads to the best between-study consistency [59] . Throughout this work , we used the rat as reference organism and mapped human and mouse genes to rat using the RGD resource [60] . We downloaded the files GENES_RAT . txt and RGD_ORTHOLOGS . txt from ftp://rgd . mcw . edu/pub/data_release/ on 2015/01/15 . This resulted in 19269 rat genes , mapped to 17459 human and 17366 mouse genes . Of these , 14078 / 15991 / 15691 are represented on the corresponding Affymetrix microarrays after the selection process described above . The complete mapping of rat genes , orthologs , selected Affymetrix probe sets and their intensities in control samples are provided in S5 Dataset . Each experiment was described using a vector of features describing gene expression changes obtained by comparing treatment and control samples . A commonly-used analysis approach involves calculating fold change values for each gene represented on the microarray ( and log-transformed to increase the normality of the distribution of fold change values ) . Since microarrays provide less reliable estimates of abundance for low-expressed genes [3 , 23] , we assembled a set of genes we describe as “rat liver-expressed genes” throughout this manuscript . Working from 20269 probe sets mapped unambiguously to one or more genes on the RG230-2 microarray , we identified a subset of 11884 probe sets with above median intensity ( calculated on each microarray ) in 10% or more of control arrays ( expressed ) or 10% of treatment arrays ( inducible ) . Where a gene is represented by multiple probe sets among the set of 11884 , we retained the highest intensity probe set as representative . This results in 9071 genes ( identified in S5 Dataset ) represented by one probe set per gene , or approximately 64% of the genes on the array . This result is in agreement with the approximately 60% of the genome estimated to be expressed in human liver [61] . Fold change values for each of the 9071 genes are described throughout this work as gene-level analysis . There are many approaches for producing estimates of mRNA abundance from Affymetrix microarrays . Perhaps the most common “legacy” approach consists of conventional ( manufacturer’s ) probe sets with MAS5 normalization , using updated probe set annotations supplied by the manufacturer . A similar approach using conventional probe sets with RMA normalization is more widely applied in the recent literature . There is a large body of literature claiming advantages for one method vs . others , yet both MAS5 and RMA on conventional probe sets perform similarly in what is perhaps the most definitive comparison from large spike-in experiments ( “MAS 5 . 0 , RMA , GCRMA yielded between 84%-87% sensitivity at a 5% FDR” ) [62] . More recently , probe sets have been redesigned by mapping underlying probes to modern genome builds ( e . g . BrainArray probe sets ) , which has been shown to have significant effect for certain genes [63] . A comprehensive assessment from the SEQC initiative[64] demonstrates increased mutual information vs . RNA-seq for a newer analysis pipeline vs . the legacy approach . We continue to use the legacy approach and RG230-2 arrays for toxicogenomics applications to facilitate the interpretation of current studies in the context of thousands of legacy experiments where re-analysis is impractical . In order to evaluate the dependence of our results on the choice of normalization and probe set definitions , we repeated analyses in this work with updated methods . As a representative “modern” approach for microarray data analysis , we selected RMA normalization and BrainArray version 19 probe sets designed against Entrez gene ( http://brainarray . mbni . med . umich . edu/Brainarray/Database/CustomCDF/19 . 0 . 0/entrezg . asp ) , and mapped them to the conventional ( i . e . manufacturer’s ) probe sets via the Entrez gene ID . Of the 9071 liver-expressed genes using conventional probe sets , 8094 have a corresponding BrainArray probe set . For a given experiment , we compare logFC values for the 8094 genes using the modern and legacy approach , and calculate the Pearson correlation . We obtained minimum / 10th percentile / median Pearson R values of 0 . 49 / 0 . 67 / 0 . 78 across the 654 DM rat liver experiments . When analyzing expression changes via modules or GSA , we find higher agreement for modules ( minimum / 10th percentile / median values of 0 . 63 / 0 . 84 / 0 . 92 ) and similar agreement for GSA ( minimum / 10th percentile / median values of 0 . 28 / 0 . 64 / 0 . 82 ) . The median level of agreement is similar to that seen when analyzing the same samples with the same method across different sites in the MAQC-I initiative ( S8 Table ) . We repeated all cross-source and cross-model comparisons with expression data from the modern processing approach . Here , we focus on three comparisons within TG because they represent trends based on hundreds of drugs ( S10 Table ) , although the same behaviors are seen for the other comparisons . For genes and modules , concordance as assessed via Pearson R or the overlap metric increases or decreases slightly when using modern processing , with increases limited to ΔR ≤ 0 . 08 and Δ overlap ≤ 4%; success at self-identification decreases slightly . The improvement is greater for GSA concordance ( ΔR ≤ 0 . 16 and Δ overlap ≤12 ) , although as observed for the other methods performance in self-identification mostly decreases ( and GSA was already the worst performing of 3 methods for self-identification when using legacy processing ) . Versions of Figs 2 , 4 and S4 created using modern processing are given in S6 , S7 and S8 Figs , respectively , and the full data is available in S2 Dataset . In summary , the results presented in this work are minimally impacted by methods used for microarray data processing . Unless otherwise noted , results in this work use MAS5-normalized gene expression results . Identification of pathways enriched among differentially expressed genes is frequently performed to simplify interpretation of results from expression profiling studies [65] . The popular GSEA algorithm [5] is perhaps the current “gold-standard” approach , and unlike classic enrichment tests that require definition of what constitutes a differentially-expressed gene ( e . g . fold change > 1 . 5 and p-value < 0 . 05 ) , GSEA and related gene set analysis ( GSA ) algorithms operate on gene-level statistics and avoid the need for arbitrary cutoffs . Here we use the PAGE algorithm [36] implemented in the Piano package since its run time is substantially less than GSEA and provides similar results [65] . The syntax used in this work is runGSA ( genes , gsc = gsc , geneSetStat = "page" , gsSizeLim = c ( 3 , 5000 ) , signifMethod = "nullDist" , adjMethod = "BH" ) . We use only one probe set per gene , selected based on intensity in controls ( see above ) . To conform with standard GSA analysis , we use all genes on the microarray , not only the 9071 liver-expressed genes . Gene sets for GSA analysis were selected from two different sources . The canonical pathway collections ( CP ) from MSigDB version 4 [5] were obtained from http://www . broadinstitute . org/gsea/downloads . jsp . The provided human Entrez geneIDs were validated and updated if necessary using the gene_info and gene_history files from NCBI ( ftp://ftp . ncbi . nlm . nih . gov/gene/DATA/ ) . The resulting human Entrez gene IDs were converted to rat gene IDs using the orthology map described above , resulting in 1320 CP gene sets ranging in size from 4 to 823 rat genes . In total , 59497 gene vs . gene set pairs were obtained involving 7572 unique rat genes . For gene ontology ( GO ) terms , we obtained the GO ontology from http://purl . obolibrary . org/obo/go/go-basic . obo , the official RGD GO to gene association for rat from http://geneontology . org/gene-associations/gene_association . rgd . gz , using all evidence codes . The latter provides identifiers for RGD and UniProtKB gene and protein identifiers . We used the above-mentioned GENES_RAT . txt file for RGD gene ID to Entrez gene ID conversion , and the UniProt mappings to Entrez gene from ftp://ftp . uniprot . org/pub/databases/uniprot/current_release/knowledgebase/idmapping/by_organism/RAT_10116_idmapping . dat . gz . We produced the full GO to gene association for the mappings in gene_association . rgd by propagating up in the GO ontology using a perl script ( “is_a” , “part_of” , “positively_regulates” , “negatively_regulates” and “regulates” associations ) . We retained GO terms of the “biological process” and “cellular component” types containing between 3 and 5000 rat genes . This resulted in 10261 GO terms and 1089835 GO term vs . gene pairs involving a total of 15408 rat genes . As done for the gene-level analysis , we described the transcriptional effects of drug treatment as a vector of Z-scores from PAGE , using one Z-score for each gene set ( the Z-score is the statistic indicating the degree of enrichment for a given gene set among the most induced or repressed genes ) . Gene sets may contain genes that are not co-expressed ( either in liver or any tissue or cell ) , and hence will not be identified as enriched among differentially expressed genes in liver experiments . In order to reduce the sparseness of this vector , we identified a subset of 1719 gene sets from the full set of 10261 GO and 1320 CP gene sets ) that have BH-adjusted p ≤ 0 . 01 in 1% of TG rat liver experiments or more ( ≥35 experiments ) . We added back 121 gene sets falling below this threshold , but pertaining to processes of relevance to drug-induced liver injury ( inflammation , oxidative stress , endoplasmic reticulum stress , ubiquitination , apoptosis , etc . ) The complete set of 1840 gene sets used for GSA analysis is provided in S6 Dataset . We used the WGCNA [37] package in Bioconductor to derived co-expression networks using the 654 DM rat liver experiments ( henceforth “training” data ) and 9071 liver-expressed genes ( i . e . a matrix with 9071 rows and 654 columns , containing log10 fold-change value from MAS5 intensities ) . We used the standard power-law plotting tool in WGCNA to set the soft-power parameter to 8 . Modules were unsigned ( i . e . can contain both induced and repressed genes; TOM = TOMsimilarity ( adjacency , TOMType = "unsigned" ) ) . The gene dendrogram was built using average-linkage hierarchical clustering ( geneTree = flashClust ( as . dist ( dissTOM ) , method = "average" ) ) , and the modules created with the dynamic branch cut algorithm ( cutreeDynamic ( dendro = geneTree , distM = 1-TOM , deepSplit = 4 , pamRespectsDendro = FALSE , minClusterSize = 5 ) ) . This yielded 354 modules containing a total of 8014 genes . An additional set of 61 merged modules were defined by grouping together 131 modules having Pearson correlation of their eigengenes ( see below ) R ≥ 0 . 8 across the training data . We describe the 61 merged + 223 unmerged ( i . e . those having max R < 0 . 8 vs . the other 353 modules ) as “base” modules . For any given experiment , each module is described by one module score , or “eigengene” , using principal components analysis ( PCA ) performed on the training data for its component genes . Prior to PCA , log10 fold change values ( logFC ) are Z-scored using the average and standard deviation of each gene’s logFC values across the training experiments . The fraction of variance of the underlying Z-scored logFC values explained by the first PC has a quasi-normal distribution , with average / standard deviation of 0 . 48 ± 0 . 09 across the 415 modules . Modules containing more genes have lower variation explained by the first PC ( Spearman ρ = -0 . 67 ) . The standard deviation of raw module scores , calculated using the 654 DM rat liver experiments , varies by module and is explained largely by the number of genes in the module ( ρ = 0 . 97 ) . In order to simplify interpretation , a final module score is obtained by dividing the raw module score by its standard deviation within the training data . As such , a module score for an experiment reflects that drug’s effect on the module measured in standard deviations across the training data . Throughout this work , we used the average absolute eigengene score ( avg . abs . EG ) as a measure of overall transcriptional activity . This quantity is calculated on the 284 base modules only in order to minimize redundancy among the modules . The WGCNA literature is dominated by sample-level analyses , i . e . human samples , cell lines , etc , where gene-level intensities are centered across samples by virtue of microarray or RNA-seq normalization algorithms . When compared to a control group , many genes have non-zero logFC . This is not surprising given the expected preponderance of xenobiotic and other responses in a dataset like DM , where doses selected were often high in order to generate histology findings in rat liver . For example , the Abcc3 subunit of the MRP drug transporter ( probe set 1369698_at ) has an average fold change of 3 . 2 . Centering the data for such genes causes experiments where logFC is near zero to take on ( usually small ) positive or negative values after subtracting average logFC across the training data . Since modules contain co-expressed genes , experiments where the component genes have logFC close to zero can appear as induced or repressed . For this reason , we center and scale for determining a gene’s weight in the module at the module building stage , but we simply scale logFC values by the standard deviation of the gene and multiply by the gene’s PCA loading when scoring experiments . The process is illustrated for an example module in S7 Dataset . Module definitions , gene weights and average / standard deviation of logFC across DM and TG are given in S8 Dataset . Preservation analysis serves to evaluate the relevance of a module in other systems [66] . We used the module Preservation function in the WGCNA package to calculate the Z-summary preservation score in TG rat liver and TG RPH . The score encompasses a number of metrics for assessing the properties of a module , and is normalized vs . the equivalent scores for modules consisting of random gene selections . We used 200 random gene selections to calculate module preservation statistics . WGCNA analysis can be regarded as producing gene sets having different weights in different systems ( or biological “contexts” ) , whereas GSA and related methods give all genes equal and non-varying weight . As such , parameters determined in one system will not be meaningful in all systems . Scoring of modules requires transforming logFC values to Z-scores using a gene’s stdev of logFC ( i . e . scaling on gene variability ) , and applying the weight of that gene in a given module . Both scaling and weights will vary across systems , giving rise to several possible choices when applying modules derived from DM rat liver to other systems: 1 ) assume that logFC stdev and gene weights from the derivation system apply in the new system , 2 ) use logFC stdev in the new system to Z-score logFC values , but use gene weights from the derivation system , or 3 ) use logFC stdev from the new system and re-calibrate the weight of genes within modules using PCA . For small datasets , like the mouse data discussed in this work , re-deriving module weights and logFC statistics would be challenging and 1 ) constitutes the only practical choice . The concordance data presented in this work uses DM liver weights for all systems , with variation in data source selected for scaling: DM rat liver for all rodent liver results , DM RPH for all RPH results , and TG HPH for HPH and HepG2 results . We investigated the impact of scaling and re-calibration post-hoc by comparing module scores for various scenarios ( S11 Table ) . The impact of using different data sources for scaling is small , in part due to high correlation of gene variability of the 9071 liver-expressed genes across sources ( S12 Table ) . Recalibration from DM liver weights to RPH or HPH weights has a larger effect , but this is markedly reduced when focusing only on modules that are preserved in the new system ( Z-summary ≥ 3 ) . This result is intuitive: the preservation score of a module measures the degree of co-expression of its member genes in the new system , and having high co-expression in both systems results in similar gene weights when calibration is performed in either system . Thus , it appears practical to apply gene weights across related systems ( in this case , systems consisting primarily of hepatocytes ) , and evaluate preservation of the module in the new system where sufficient perturbation data allow . Similarly , we investigated the impact of using RMA-normalized DM liver data ( with conventional probe sets ) for module calibration and found minimal effects on scoring of experiments ( minimum / 10th percentile / median Pearson R values of 0 . 82 / 0 . 94 / 0 . 98 comparing MAS5 and RMA scores for 654 DM liver experiments ) . When WGCNA analysis was performed at the onset of this work , we selected the probe set with the highest standard deviation of logFC values across the rat liver DM dataset as the representative for a given gene ( i . e . the most variable probe set for the gene ) . This was subsequent to application of the intensity-based filter: a probe set still needed to be above median intensity in 10% of control or treatment samples . The combination of intensity and variability-based selection was impractical for systems where we have few samples ( i . e . 10% of samples might represent a few samples for small datasets ) . When using DM rat liver data , the probe set selected for the 9071 liver expressed genes based on maximal intensity differs from that selected as most variable for 1670 genes . For application of the rat liver-derived modules to other systems , we adopted the recommendation from [59] in selecting as representative probe set that with maximal intensity . The eigengene scores used in this work use the probe set selected by variability for scoring rat liver and RPH , and intensity-selected probe sets for mouse and human samples . We re-calibrated the original DM rat liver modules using the highest intensity probe set for these 1670 genes , to obtain gene weights that correspond to the higher intensity probe set . Comparison of module scores using the original variability-based and intensity-based selection across the 654 rat liver DM experiments gives minimum / 10th percentile / median Pearson R values of 0 . 83/0 . 94/0 . 98 . The variability-based and intensity-based probe set selections , along with weights required for module scoring with either selection are provided in S8 Dataset . Overall , module eigengenes are robust to changes in genes / probe sets selected . We re-derived the modules on the DM rat liver data using the complete set of 14078 genes , selecting the highest intensity probe set for each gene and applying no intensity cutoff to eliminate low expression genes , obtaining 381 modules that contain 8639 genes in total ( vs . the 8014 contained in modules on our original build ) . When calculating the pairwise correlation between original and new modules across the 654 DM rat liver experiments , 374 of 415 original modules have Pearson R ≥0 . 7 vs . a new module , and 308 have Pearson R ≥ 0 . 8 . Percent overlap is a qualitative metric popularized by the MAQC initiative[3 , 23 , 24] which simply reports what percentage of the top-ranked features overlap between two experiments . We illustrate its implementation in this work using the 9071 liver-expressed genes , with equivalent applications to GSA or modules at the same 5% threshold . For each experiment , the top 5% of genes are identified , with no consideration of direction ( induced or repressed ) , giving 454 genes . When comparing two experiments , we identify the number of genes in common between both lists of 454 genes and changing in the same direction in both experiments . This value is divided by 454 and reported as a percentage . We examined the distribution of average absolute eigengene scores ( avg . abs . EG ) for 78 , 226 pairs of TG rat liver experiments and 10 , 008 TG RPH experiments , and divided the range into 7 intervals: <0 . 2 , 0 . 2–0 . 3 , 0 . 3–0 . 4 , 0 . 4–0 . 5 , 0 . 5–0 . 6 , 0 . 6–0 . 8 and ≥ 0 . 8 . For each of 3 sets of gene expression features ( genes , gene sets and modules ) and two metrics ( Pearson R and percent overlap ) , we generated 1000 random pairs of experiments involving different drugs falling into each avg . abs . EG range . When evaluating the concordance for a pair of experiments involving the same drug , we report the proportion of the 1000 random pairs with equal or greater concordance on the same feature and metric and falling in the same avg . abs . EG range . When defining success at ‘self-identification’ , we require fewer than 5% of random pairs to exceed the level of concordance seen for the pair in question . Several metrics are available for quantifying the similarity of two gene expression profiles . In this work we have selected Pearson R and percent overlap because we consider them to be intuitive metrics when comparing two conditions ( plotting genes’ logFC via scatter plots , for example ) . The similarity of expression profiles can be quantified with the Euclidian distance , and this represents a common choice when performing principal components analysis and clustering on gene expression profiles [64] . Identical gene expression profiles will have a Euclidian distance of 0 and Pearson correlation of 1 . We calculated the Euclidian distances between 78 , 226 pairs of rat liver experiments and 10 , 008 pairs of RPH experiments from TG , and compared them to the Pearson distance ( 1 –Pearson R ) . Euclidian and Pearson distances are uncorrelated for genes and modules across both systems ( R < 0 . 17 ) , and modestly correlated for GSA ( R = 0 . 53 and 0 . 63 for rat liver and RPH , respectively ) . Euclidian distance effectively regards a common absence of differentially expressed features as evidence of similarity , giving equal importance to the usually far larger number of non-differentially expressed features compared to Pearson R or overlap metrics . We evaluated the performance of Euclidian distance in self-identification for all comparisons depicted in Figs 2 , 4 and S4 . Compared to the Pearson R ( S9 Fig ) and overlap metrics ( data in S2 Dataset ) , Euclidian distance performs similarly for GSA and notably worse for genes and modules . Due to its lower performance and less intuitive characteristics ( “your molecule is similar to drug X because neither of have a significant effect on the liver transcriptome” ) , we do not advocate its application for CMap-type applications . Results for all comparisons , using both legacy and modern array normalization methods , are provided in S2 Dataset . | Gene expression studies in model systems are widely used for understanding the mechanism of drugs and other perturbations in biological systems . Other researchers have examined the reproducibility of microarray studies between laboratories , or comparing microarrays and/or RNA sequencing . However , no large scale studies have compared results from protocols which differ in minor details , or results generated in vivo vs . in vitro culture systems thought to serve as useful models . The rat liver is by far the most extensively studied model evaluating effects of drugs and other perturbations , and existing data allowed us to assess the level of concordance between rat liver and rat primary hepatocytes cultured in collagen-coated plates ( i . e . “flat” culture ) for hundreds of drugs . We found that the mouse liver serves as a better model of the rat liver than do rat primary hepatocytes , even after allowing for differences due to pharmacokinetics . The low concordance observed between rat liver and rat hepatocytes suggests that validating the utility of ‘omics data generated on emerging cell culture approaches ( e . g . “organ-on-a-chip” , 3D-printed tissues ) using rat cells and comparison to the rat liver may be necessary in order to gain confidence these approaches substantially improve on traditional culture models of human cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"liver",
"vertebrates",
"animals",
"mammals",
"dna",
"transcription",
"animal",
"models",
"model",
"organisms",
"pharmaceutics",
"genome",
"analysis",
"bioassays",
"and",
"physiological",
"analysis",
"research",
"and",
"analysis",
"methods",
"genomics",
"animal",
"cells",
"gene",
"expression",
"hepatocytes",
"gene",
"ontologies",
"microarrays",
"rodents",
"cell",
"biology",
"anatomy",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"drug",
"therapy",
"amniotes",
"computational",
"biology",
"organisms",
"rats"
] | 2016 | Assessing Concordance of Drug-Induced Transcriptional Response in Rodent Liver and Cultured Hepatocytes |
Genome-wide association studies have successfully identified thousands of loci for a range of human complex traits and diseases . The proportion of phenotypic variance explained by significant associations is , however , limited . Given the same dense SNP panels , mixed model analyses capture a greater proportion of phenotypic variance than single SNP analyses but the total is generally still less than the genetic variance estimated from pedigree studies . Combining information from pedigree relationships and SNPs , we examined 16 complex anthropometric and cardiometabolic traits in a Scottish family-based cohort comprising up to 20 , 000 individuals genotyped for ~520 , 000 common autosomal SNPs . The inclusion of related individuals provides the opportunity to also estimate the genetic variance associated with pedigree as well as the effects of common family environment . Trait variation was partitioned into SNP-associated and pedigree-associated genetic variation , shared nuclear family environment , shared couple ( partner ) environment and shared full-sibling environment . Results demonstrate that trait heritabilities vary widely but , on average across traits , SNP-associated and pedigree-associated genetic effects each explain around half the genetic variance . For most traits the recently-shared environment of couples is also significant , accounting for ~11% of the phenotypic variance on average . On the other hand , the environment shared largely in the past by members of a nuclear family or by full-siblings , has a more limited impact . Our findings point to appropriate models to use in future studies as pedigree-associated genetic effects and couple environmental effects have seldom been taken into account in genotype-based analyses . Appropriate description of the trait variation could help understand causes of intra-individual variation and in the detection of contributing loci and environmental factors .
Phenotypic variation for a quantitative trait is attributable to the summed effects of genetic and environmental influences together with any covariances and interactions . The proportion of phenotypic variance contributed by genetic variation is termed the heritability ( h2 ) [1] . The heritability scales the influence of genetic and environmental factors on phenotypic variation . This provides us with insights into the genetic and environmental architecture of human complex traits and our potential ability to dissect out loci associated with trait variation and is also useful for the prediction of heritable disease risk [2 , 3] . As a consequence , such knowledge is of potential value for clinical diagnosis , therapy , prevention and prognosis [4] . Therefore , obtaining unbiased estimates of variation caused by different factors and the heritability of traits relevant to health and disease processes is important . A classic approach to gauging the heritability in humans is by comparing the observed phenotypic similarity to the expected genetic resemblance between relatives inferred from family pedigrees [5] . This method evaluates the pedigree based heritability ( hped2 ) indirectly without requiring information on the inheritance of individual loci and thus , is quite practical and still widely-used in twin , family and other pedigree studies [6 , 7] . Note that , hped2 is often considered to be an estimate of the true heritability h2 . Genome-wide association studies ( GWAS ) , on the contrary , identify causal loci through their association with recorded genetic markers and then aggregate the proportion of variance explained by statistically-significant variants [8 , 9] , which is sometimes referred to as the “GWAS heritability” ( hGWAS2 ) . Each approach has its limitations and drawbacks . Pedigree studies require genealogical information from known relatives to deduce their expected genetic resemblance and hped2 may be biased due to the factors shared among relatives ( including dominance , epistasis , common environment , genetic-by-environment correlation and genetic-by- environment interaction ) if such effects are present and the available pedigree structure does not allow these to be accounted for in the analysis [10–12] . Although GWAS have been very successful at discovering novel loci for a range of polygenic disease and complex traits , they have been less successful at capturing the full extent of known trait genetic variance [11 , 12] . This is probably because of their failure to detect particular types of variants such as common variants with small effects , rare variants , copy number variants and structural variants , as a consequence of inadequate sample size , genotyping platform design and analyses used , together with the stringent statistical tests applied [10 , 13 , 14] . As a result , there usually is a substantial gap between the estimates of hped2 and hGWAS2 , often termed the “missing heritability” [11 , 15] . Recently , Yang et al . [16 , 17] have championed an approach , known as GREML [18] , to estimate the amount of trait variance explained by SNPs . The estimation of the SNP ( or genomic ) heritability ( hg2 ) , which refers to the additive genetic effects captured by genotyped SNPs , utilises a matrix comprising realised genetic relationships inferred from genomic marker data originally gathered for GWAS ( known as genomic relationship matrix or GRM ) [16 , 17] . The hg2 estimate from this approach , when estimated using unrelated individuals , lies between the hped2 and hGWAS2 estimates , and has been considered as a lower limit for the former and an upper limit for the latter [11 , 12] . As an example , for height , hGWAS2 , hg2 and hped2 from three different studies are 0 . 10 , 0 . 45 and 0 . 80 respectively [5 , 8 , 17] . This suggests that a substantial proportion of the genetic contribution to trait variation is SNP-associated and hence contributes to hg2 but not all this variation is detected by current GWAS , probably due to a combination of insufficient sample size and stringent significant thresholds employed . The difference between hg2 and hped2 may be largely due to trait associated variants not in linkage disequilibrium ( LD ) with genotyped SNPs , such as rare variants , copy number variants ( CNV ) and other structural variants as mentioned above . Variation associated with such effects is captured by hped2 due to strong LD in relatives [19] . Recent studies have started dissecting the heritable component of variation and other components shared among relatives by studying more complex populations made-up of both unrelated individuals and extended pedigrees [11 , 12 , 19] . For instance , Zaitlen et al . [12] have demonstrated that simultaneously including in a GREML analysis a GRM and a modified GRM ( in which entries smaller than a certain threshold in the GRM are set to zero ) can be used to jointly estimate SNP-associated and total heritabilities in the presence of relatives . We also note that shared environment may be an important contributor to heritability inflation when close relatives are included in analysis . In this study , we use data from a single homogeneous cohort consisting of approximately 20 , 000 adults with varying degrees of relationships sampled from Scotland . The individuals have data on over 520 , 000 SNPs distributed across the autosomes . The dense marker information together with extended genealogical information allows us to partition the phenotypic variance and explore the genetic and environmental effects shared among related individuals ( both biological relatives and couples ) . We analyse eight anthropometric traits , comprising height , weight , fat , body mass index ( BMI ) , hips , waist , waist-to-hips ratio ( WHR ) and a body shape index ( ABSI ) [20] and eight cardiometabolic traits , comprising levels of creatinine , urea , total cholesterol ( TC ) and high density lipoprotein ( HDL ) in serum , level of glucose in blood , systolic blood pressure ( SBP ) , diastolic blood pressure ( DBP ) and heart rate ( HR ) . In our work , we implement alternative models to estimate effects that might contribute to the variation in the 16 traits analysed . Results show that , with these data , we can separate total genetic variation into SNP-associated and pedigree-associated genetic influences . We also observe that past family environment and shared full-sibling environment generally have a limited impact on trait variation , whereas the effect in couples of living in the current ( shared ) environment is always important in our data .
We conducted variance component analyses to dissect the phenotypic variation for traits recorded in the Generation Scotland: Scottish Family Health Study ( GS:SFHS ) cohort [21] into genetic and environmental factors . Analyses utilised a mixed-model approach implemented in a restricted maximum likelihood ( REML ) framework using the GCTA software [16] . The population was divided into two tranches of approximately equal size and genotyped in two stages . All initial analyses were performed with the first 10 , 000 genotyped individuals , ( named GS10K ) . GS10K comprised small nuclear families ( largely two parents and two offspring ) together with unrelated individuals , although inevitably there were second degree and more distant relationships included . The second tranche completed the genotyping of the rest of the population ( another 10 , 000 individuals ) including further relatives in incomplete families ( e . g . missing samples from parents and additional siblings , as well as other relationships ) , resulting particularly in a proportional increase in the number of second and third degree relationships ( Table 1 ) . To confirm results obtained from GS10K , some of the analyses were repeated in the whole 20 , 000 individual sample ( named GS20K ) . We first explored the extent to which estimates of hg2 were inflated by the inclusion of relatives . We subsequently analysed our data allowing trait variation to be potentially influenced by both genetic and environmental effects . We assumed that the genetic effects comprised additive genetic effects associated with genotyped SNPs ( hg2 ) and additional additive genetic effects associated with pedigree but not with genotyped SNPs ( hkin2 ) , and we assumed that the environmental effects potentially comprised nuclear family effects ( ef2 ) common to both parents and offspring , full-sibling effects ( es2 ) common to just siblings and couple effects ( ec2 ) common to just the members of a couple ( Fig 1 ) . The total heritability , termed hgkin2 in this manuscript , referred to as hIBS>t*2 in Zaitlen et al . [12] and comparable to hped2 from traditional pedigree studies , was estimated as the sum of hg2 and hkin2 for each model . To allow estimation of the influence of each effect , we generated five design matrices: GRMg , GRMkin , ERMFamily , ERMSib and ERMCouple respectively , where GRM refers to genomic relationship matrices and ERM refers to environmental relationship matrices . For brevity , we named different alternative models using abbreviations according to first subscript letter of the effects examined . We coded ‘G’ for GRMg , ‘K’ for GRMkin , ‘F’ for ERMFamily , ‘S’ for ERMSib and ‘C’ for ERMCouple–e . g . model ‘GKC’ = GRMg + GRMkin + ERMCouple . All models included a residual matrix ( allowing effects specific to an individual that were not shared with any other member of the population ) . We identified the most appropriate model for each trait by a stepwise model selection process via removing non-significant components from the full model based on a Wald test of their estimated effect and a likelihood ratio test ( LRT ) , and we estimated the effects of significant factors using the selected models in GS10K . We repeated the model selection and corresponding variance component analyses in GS20K to identify differences resulting from analysing a more complex population structure , encompassing a larger proportion of close relationships . More details about traits , matrices and models are given in Material and Methods and S1 Table and S2 Table . In the main manuscript , we only list results for the final models identified by the model selection procedure and the full model , but a comprehensive list of estimates obtained for the different effects for each trait and each model is available in S3 Table and S4 Table . Model robustness and the effectiveness of the model selection were tested using simulated data based on GS10K . We conducted a simulation study using real genotype and pedigree information from GS10K to evaluate the robustness of our models . To make computation feasible , we mainly focused on data simulated under the simplest and most complex models ( models ‘G’ , ‘K’ , ‘F’ , ‘S’ , ‘C’ , ‘GK’ , ‘GF’ and ‘GKFSC’ ) and those representing the commonest conclusions of model selection in analyses of the real GS10K data ( models ‘GF’ , ‘GFS’ , ‘GKC’ and ‘GKSC’ ) . S5 Table shows the simulated and observed values for each parameter as well as the model we used for analyses in different scenarios . In the first scenario , we examined the performance of our models ( models ‘G’ , ‘K’ , ‘F’ , ‘S’ and ‘C’ ) when simulated phenotypes were only contributed by one of the five corresponding effects plus residual variation . Under these models ( S5 Table ) , the mean of overall estimates per parameter was very close to its simulated value , indicating that our design matrices GRMg , GRMkin , ERMFamily , ERMCouple and ERMSib worked well in simple models and were able to capture their corresponding effects even when the simulated variance associated with an effect was low ( ≤ 3% ) . In the second scenario , we evaluated the performance of our models ( models ‘GK’ and ‘GF’ ) when the simulated phenotypes were determined by SNP-associated genetic effects and one of the familial effects ( either pedigree-associated genetics or nuclear family environment ) plus residual variation . Results ( S5 Table ) indicate that , in cohort with familial structure , failure to account for or inaccurate modelling of familial effects ( i . e . when models used were inconsistent with phenotypic contributors ) would result in upward bias for hg2 in the presence of relatives . However , this upward bias due to the confounding familial factors could be eliminated by either excluding nominally related individuals or using the appropriate models for analysis . The former method removes the ability to estimate the familial effects as well as reducing the sample size , whereas using the appropriate models , estimates obtained were very close to their parameter settings and gave a good idea of the magnitude and approximate values of SNP and familial effects as well as the total proportion of variance explained by additive genetics ( hgkin2=hg2+hkin2 ) , despite the fact that the means of estimates of hg2 , hkin2 and ef2 were usually significantly different from the original parameter settings . In the third scenario , we inspected the performance of the full model ‘GKFSC’ and models selected from analyses of real phenotypes in GS10K other than ‘GF’ ( models ‘GFS’ , ‘GKC’ and ‘GKSC’ ) . Results ( S5 Table ) demonstrate that all models were robust in terms of the mean of overall estimates per parameter being either unbiased or very close to original settings . Fig 2 summarizes the main results from these simulations , showing the overall performance of our design matrices from simple models to complex models . The median of estimates for each component was unbiased across simple and complex models , however , the estimates for hkin2 , ef2 and ec2 were quite variable in the full model , probably due to limitations imposed by the data structure . All of the above verify the robustness of our models . Although we confirmed that our models were robust ( S5 Table and Fig 2 ) , the potentially high correlation between ERMFamily matrix and combined ERMCouple and GRMkin matrices may make it challenging to jointly estimate hkin2 , ef2 and ec2 accurately in our sample as the standard errors for those parameter estimates obtained from the full model were high ( S4 Table ) . Thus the most challenging part of our study may be to precisely dissect pedigree-associated genetic effects , shared nuclear family environment and shared couple environment . Therefore , we performed model selection using simulated data to test our model selection procedure where simulated phenotypes were contributed by moderate SNP-associated genetic effects and low sibling environmental effects plus a ) moderate nuclear family environmental effects but low pedigree-associated genetic effects and couple environmental effects; b ) low nuclear family environmental effects but moderate pedigree-associated genetic effects and couple environmental effects; or c ) moderate nuclear family environmental effects , pedigree-associated genetic effects and couple environmental effects . All scenarios included residual variation . S6 Table shows the parameter settings and the summary of model selection procedure performance for these scenarios . We expected that our model selection procedure was able to identify SNP genetics ( GRMg ) and nuclear family environment ( ERMFamily ) or SNP and pedigree genetics ( GRMkin ) and couple environment ( ERMCouple ) or SNP and pedigree genetics and nuclear family and couple environment accordingly , since they were the major factors in each corresponding scenario . As results demonstrated , in all situations our model selection procedure generally ( ≥80% ) selected the appropriate model which contains all major components of phenotypic variation . The remaining times in the first two of these scenarios , pedigree-associated genetic effects or those plus shared couple environment were selected instead of nuclear family environmental effects or vice versa , and in the remaining two replicates in the third of these scenarios we missed pedigree-associated genetic effects . In addition , our model selection never fully detected all minor contributions to the phenotype in the first two of these scenarios when the minor effects were too small ( e . g . effects contribute to ≤5% of the phenotypic variance ) . Both issues identified above ( ~20% chance of selecting inappropriate models and failure to identify all minor effects ) are likely to have been due to limitations in the data structure of GS10K , which provides too few of the appropriate relationships for corresponding effects ( pedigree-associated genetics , nuclear family , sibling and couple environment ) to resolve correlations between parameters and detect minor effects . These limitations have been greatly ameliorated in the GS20K data . We also conducted variance component analyses using the final selected model for each replicate ( S6 Table ) . For those replicates that had appropriate models after model selection , the estimates of factors that remained in the models were usually close to , and not significantly different from , their simulated values , indicating that the results from selected models were reliable . More details about simulation study can be found in S1 Text , S5 Table and S6 Table . In the first analyses of the real data , we looked for evidence of familial effects ( either pedigree-associated genetics or nuclear family environment ) in our cohort . As shown by simulation ( S5 Table ) , if there were any familial effects , we should obtain inflated estimates of hg2 when we conducted variance component analyses using model ‘G’ in the presence of relatives , compared to the estimates of hg2 given from the unrelated subpopulation . GS10K consists of nearly 10 , 000 genotyped individuals with multiple degrees of relationship , which allows us to explore the impact of familial effects on hg2 estimation in this cohort . Table 1 shows the population structure of genotyped individuals in GS10K . The degree of relationship between two individuals was identified according to an approximate range of the expected pair-wise relatedness ( r ) , which was from 0 . 5i-0 . 5 to 0 . 5i+0 . 5 for ith degree relatives ( e . g . pairs of individuals with relatedness from 0 . 354 to 0 . 707 were considered as 1st degree relatives ) . With these criteria , GS10K consisted of more than 3 , 500 pairs of 1st degree relatives , around 450 pairs of 2nd and 500 pairs of 3rd degree relatives , but the majority of pairs of individuals ( over 99 . 9% ) were genetically unrelated ( more distant than 5th degree relatives , r ≤ 0 . 022 ) . In total , there were around 6 , 600 unrelated individuals ( defined using the criteria described above ) in GS10K . We estimated hg2 for each trait using model ‘G’ for subpopulations of GS10K made-up of individuals with different degrees of relatedness ( using the upper bound of the expected relatedness of each category as GRM cut-off points in GCTA ) . Fig 3 shows how hg2 estimates for height , BMI and HDL changed as we progressively included more closely related individuals in the relationship matrix . Results for the remaining traits are shown in S3 Table . In general , hg2 estimates were stable as we gradually added more closely related individuals in the analyses until the inclusion of 1st degree relatives that resulted in inflation of the estimates ( Fig 3 and S3 Table ) . Based on our results , hg2 was overestimated only when 1st degree relatives were included . For glucose and DBP , the hg2 estimates did not appear inflated after 1st degree relatives were included , suggesting that these traits were not affected by familial effects ( S3 Table ) . The increase in hg2 estimates resulting from the inclusion of 1st degree relatives provided evidence of familial variation in our cohort . However , it is not clear whether these familial effects are due to pedigree-associated genetic effects or shared nuclear family environment or both because either of them has the ability to inflate hg2 estimates ( this was also observed in the simulation data: S5 Table: scenario ii ) . Therefore , we attempted to tease out this familial variance from the total phenotypic variance and dissect the familial variation as well as the remaining trait variation further using the full model ‘GKFSC’ and the stepwise selection procedure to define a final model containing the most important effects contributing to trait variation . Table 2 shows the results for final models selected from stepwise model selection strategies and for the proportions of total phenotypic variance explained by different effects using final models , as well as for those obtained using the full model . The mean estimates for hg2 , hkin2 , ef2 , es2 and ec2 across all traits in the full model were 0 . 18 , 0 . 22 , 0 . 03 , 0 . 03 and 0 . 11 , respectively . However , the majority of estimates for parameters other than hg2 obtained using the full model were not significantly different from zero according to either the Wald test or LRT performed and had large standard errors in general . These results suggest that the full model ‘GKFSC’ may suffer from the inclusion of correlated factors , as foreseen in the simulation study , probably due to a low number of different types of pairwise relationship in GS10K . Therefore , we utilised a model selection procedure designed to provide more precise estimates of the parameters retained in a more robust and parsimonious final model , where the least significant effects are removed from the model . More details about the selection procedure are given in Material and Methods . We have demonstrated the effectiveness of our model selection procedure by simulation in the previous section and S6 Table . As shown in Table 2 , SNP-associated genetic effects ( represented by GRMg ) were retained in the final models for all 16 traits , indicating that all traits examined here are heritable . Regarding variation associated with families , pedigree-associated genetic effects ( represented by GRMkin ) and nuclear family environmental effects ( represented by ERMFamily ) were retained in the final models for 10 and 4 out of 16 traits respectively . However , in GS10K , the data structure did not allow for both familial effects to be retained together in the final models for any trait . Additionally , the final models for glucose and DBP included neither GRMkin nor ERMFamily , which is consistent with the previous conclusion derived from S3 Table , suggesting that familial effects may be limited for these traits . The additional environmental influences of couple environmental effects ( represented by ERMCouple ) were retained in the final models for 12 out of 16 traits and sibling environmental effects ( represented by ERMSib ) only remained for creatinine and TC . Although the final model varied between traits , the model ‘GKC’ was most often selected ( 9 out of 16 traits ) in the model selection procedure in GS10K . Therefore , this suggests that the common environment shared by couples , SNP-associated and pedigree-associated genetic effects are important for the control of a large proportion of the human complex traits we examined , while the shared family and full-sibling environment have a more limited impact SNP-associated genetic effects ( GRMg ) in the final models provided estimates of hg2 ranging between 0 . 10 and 0 . 30 with a mean of 0 . 19 for the 15 traits , excepting height for which nearly half of its phenotypic variation ( 0 . 47 ) was SNP-associated . For the 10 traits that retained pedigree-associated genetic effects ( GRMkin ) in the final models , the estimates of hkin2 ranged from 0 . 13 to 0 . 36 with a mean of 0 . 26 , except for creatinine for which nearly half of its phenotypic variation ( 0 . 45 ) was pedigree-associated . For the 10 traits that retained both GRMg and GRMkin in the final models , the estimates of hkin2 accounted for 56% of the total heritability ( hgkin2=hg2+hkin2 ) . Regarding nuclear family environmental effects , the estimates of ef2 for 4 traits that retained ERMFamily in the final models were of 18% for anthropometric and of 10% for cardiometabolic traits . Creatinine and TC were the only two traits for which the common sibling environment ( ERMSib ) was kept in the final models , and es2 contributed 7% and 12% of their phenotypic variance respectively . For those 12 traits that demonstrated evidence of couple effects ( i . e . retained ERMCouple in the final models ) , ec2 accounted for 13 . 5% of the phenotypic variance on average ( of 15% for anthropometric traits and of 11% for cardiometabolic traits ) . Compared to the results from the full model in Table 2 , using the selected final models provided similar but more precise ( i . e . with smaller standard errors ) parameter estimates . Therefore , whereas the full models gave a general picture of the important components in the architecture of the traits , the final selected models provided a parsimonious model with more precise estimates of the most important effects . We added an extra 10 , 000 genotyped and phenotyped individuals from the same population , providing 20 , 000 individuals in total , in order to confirm and build upon the results of the model selection in a more complex data set . The difference in sample sizes and numbers of different relationships between GS10K and GS20K is shown in Table 1 . The extra 10 , 000 genotyped individuals in GS20K consisted mainly of the relatives of those already genotyped in GS10K , which substantially increased the proportion of 2nd and 3rd degree and sibling relationships in GS20K . We repeated the model selection procedure and corresponding variance component analyses using selected models in GS20K to identify changes resulting from the increased complexity and sample size of the population . Results for model selection and variance component analyses using the final selected model as well as the full model are shown in Table 3 . In general , the parameter estimates obtained from the full model in GS20K were similar to those obtained from the full model in GS10K but the number of non-significant estimates were much lower due to smaller standard errors . Note that standard errors of estimates are not only reduced using GS20K , but , unlike results from GS10K in Table 2 , are also similar between full and reduced models , suggesting the change is due to improved structure of the data to separate effects as well as increased sample size . The final models selected from model selection in GS20K were generally similar to those in GS10K , but , owing to the presence of more nuclear family members and siblings in GS20K , we now had better power to detect the past environmental effects ( either nuclear family environment or sibling environment ) , although the estimated effects were usually small . Moreover , due to an increased number and higher proportion of 2nd and 3rd degree relatives , we had better resolution for familial effects in GS20K . Pedigree-associated genetics and nuclear family environment were now separable and the data structure in GS20K can provide sufficient evidence for both types of familial effects . For weight , urea , TC and HR , familial effects switched from nuclear family environment in GS10K to pedigree-associated genetics or pedigree-associated genetics plus nuclear family environment in GS20K . However , as in GS10K ( Table 2 and S3 Table ) , there was still no evidence of either genetic or environmental familial effects for glucose and DBP in GS20K . The results from final selected models in GS20K are summarized in Fig 4 . The heritability estimate is nearly 90% , 60% and 60% for height , creatinine and HDL respectively , and for the remaining anthropometric and cardiometabolic traits , it ranges from 30%-50% and 20–30% for the two types of trait , respectively ( Fig 4B ) . Although the proportion of genetic variance explained by SNP-associated and pedigree-associated genetic effects varies across traits , each genetic effect explains around 50% of the genetic variance on average ( Fig 4C ) . In GS20K , the most commonly selected model was ‘GKSC’ ( 10 out of 16 times , Fig 4A and Table 3 ) . SNP-associated genetic effects , pedigree-associated genetic effects , sibling environment and couple environment appeared in the final models for 16 , 14 , 12 and 16 out of 16 times respectively and the means of estimates for hg2 , hkin2 , es2 and ec2 for traits which retained corresponding matrices ( GRMg , GRMkin , ERMSib and ERMCouple respectively ) in the final models were of 0 . 20 , 0 . 23 , 0 . 05 and 0 . 11 respectively ( Fig 4A and Table 3 ) . For the nuclear family environment , the mean of estimates for ef2 for 4 traits which retained ERMFamily in final models was of 0 . 04 ( Fig 4A and Table 3 ) . On average across traits , our environmental matrices and the final selected models retained through our model selection procedure could explain ~16% and ~56% of the total phenotypic variance respectively ( Fig 4B ) . The major change in GS20K compared to GS10K is the significant evidence of effects of the sibling environment , particularly for cardiometabolic traits , resulting from the higher proportion of sibling relationships in GS20K ( more than 12 times compared to GS10K , Table 1 ) . However , the sibling effects were only 5% on average and were still relatively low compared to genetic effects and couple environment . Therefore , despite the change in population structure in GS20K , the major components for anthropometric and cardiometabolic traits were SNP-associated and pedigree-associated genetic effects and couple environment as they were in GS10K ( Table 2 ) .
The aim of this study was to better understand the architecture of human complex traits by dissecting phenotypic variation into SNP-associated additive genetic variation ( hg2 ) , pedigree-associated genetic variation ( hkin2 ) and environmental influences of common environment shared by nuclear family members ( ef2 ) , full-siblings ( es2 ) and couples ( ec2 ) . We generated five design matrices GRMg , GRMkin , ERMFamily , ERMSib and ERMCouple to describe the five effects and we examined 16 human complex traits using genome-wide genotype data and genealogical information in the Generation Scotland: Scottish Family Health study ( GS:SFHS ) comprising samples from up to 20 , 000 individuals . The results of these analyses suggest that SNP-associated genetic effects , pedigree-associated genetic effects and current environment shared by couples were the major contributors to phenotypic variation for anthropometric and cardiometabolic traits . Past environmental influences , such as shared sibling environment or nuclear family environment , made relatively small or undetectable contributions to trait variation ( Table 2 and Table 3 ) . The relative importance of a couple or spousal effect for most traits was also noted by Liu et al . [22] , in analyses based only on pedigree relationships , although they did not find a significant spousal effect for cholesterol , HDL or glucose for which a significant couple effect was detected in this study . Considering the low number of non-zero off-diagonal entries in ERMCouple ( 1 , 283 or 1 , 767 pairs in GS10K or GS20K ) , the signal of couple effects was quite strong . We did observe significant phenotypic correlation between couple pairs for almost all traits in our data ( S7 Table ) . For some traits this presumably represents current shared environment due to cohabitation , such as living habits and diet . For traits related to obesity , it is reasonable that current environmental effects are more important than past environmental effects since traits like BMI , fat , HDL and blood pressure are potentially influenced by recent food intake , exercise and medical treatment . It should be noted that in our sample participants have an average age of ~50 years and individuals currently sharing a common household environment will largely be couples , whereas most individuals involved in sibling and parent-offspring relationships will no longer be cohabiting at the point when the data were recorded . It has been previously reported in obesity studies that common childhood environment only affects individuals in their mid-childhood but the influence does not last past adolescence [23 , 24] . Therefore , although the impacts of nuclear family or sibling environmental effects on the 16 traits we examined were relatively small , family and sibling environmental effects could be more important in younger cohorts and might be of greater importance for other complex traits and diseases where long-term environment may have an influence on a phenotype that is relatively stable over time . For some traits , the most obvious example being height , couple effects may also , in part or completely , reflect assortative mating . A study by Keller et al . has shown that h2 estimate for height would be 13% higher with assortative mating than it would have been under random mating [23] . If there was assortative mating for any of the traits which retained ERMCouple in final models but we modelled the couple correlation as an environmental effect , we would expect to obtain biased ec2 estimates . Moreover , modelling assortative mating as an environmental effect removes variance from the residual ( “error” ) variance . We therefore might obtain an inflated hg2 estimate if we have not taken assortative matting into account and reduce the residual variance as a consequence of modelling assortative matting as an environmental effect . In addition , assortative mating will have consequences for our interpretation of GWAS results as the combined effect of detected loci on the trait variance will be greater than the sum of the effects of the individual loci due to the positive correlations between loci . However , except for height , where the phenotype will be largely fixed by the time of marriage , for most traits it is difficult to determine whether assortative mating and/or shared environment are responsible for observed phenotypic correlations between couples . Shared sibling environment was undetected for most of the traits in GS10K ( Table 2 ) , whereas there was significant evidence of it for many traits in GS20K ( Table 3 ) , indicating that the detection power of sibling environment benefits from the increase in number and proportion of sibling relationships ( Table 1 ) . Sibling effects , where detected , explained 5% , on average , of the trait variation . Estimated sibling effects may be inflated by non-additive genetics , ( i . e . dominance and epistasis ) . As sibling effects only capture a fraction of the non-additive variation , the actual variation contributed by non-additive genetics might potentially be large and would merit further study . Our analyses split the genetic variation approximately equally on average across traits between that which was associated with SNPs ( hg2 ) and that which was associated with pedigree ( hkin2 ) . A plausible interpretation for the division of genetic effects into hg2 and hkin2 is that hg2 is able to explain the genetic variation attributed by common variants inherited from distant ancestors that are in LD at the population level and are well captured due to association with genotyped SNPs [12] . On the other hand , hkin2 accounts for the genetic variation due to rare variants , CNVs and other structural variation , etc . that cluster in specific families and are captured due to strong linkage in high-order pedigrees but are not in population-wide LD with common SNPs . We compared hg2 and hgkin2 ( calculated as hg2+hkin2 ) estimates obtained in final models from model selection in GS20K to two relevant publications from Zaitlen et al . [12] and Vattikuti et al . [19] that also explored the influence of including relatives on h2 estimation in family-based studies and compared hgkin2 estimates obtained in final models in GS20K to published twin studies [6 , 24–31] . Comparisons are shown in Table 4 . When comparing with two family-based GREML studies ( Table 4 ) , our hg2 and hgkin2 estimates from final models are generally higher than published relevant results , except for the hg2 estimate for SBP and the hgkin2 estimates for glucose and SBP . When comparing with twin studies ( Table 4 ) , our hgkin2 estimates for all anthropometric traits , urea , TC and HDL given by final selected models in GS20K are reasonably close to reported hped2 estimates , which suggests little missing heritability . Hence , our results provide no evidence that heritabilities given by previous twin studies were inflated for these traits . For glucose , SBP , DBP and HR , however , our hgkin2 estimates are significantly lower than previously published estimates of hped2 , whereas for creatinine , hgkin2 is significantly larger . To validate the analytical approach used in this study and to evaluate model robustness , we conducted a detailed simulation study using real genotype and pedigree information obtained from GS10K . The simulation results confirmed that our models were generally robust ( S5 Table ) . However , the inevitable correlations between our design matrices can , under some circumstances , make it challenging to partition variance for correlated factors in variance component analyses and accurately discriminate between competing models in model selection . Nonetheless , any influence of inaccurately partitioning variance among correlated matrices was relatively limited and our models were always able to provide us with a good idea of the magnitude of corresponding effects as the mean estimate for each parameter was always very close the simulated settings when the model used for analysis matched the simulated sources of trait variation . The effectiveness of the model selection procedure was also validated using the simulated data with the model selection procedure often ( ≥80% ) resulting in models containing all major phenotype components ( S6 Table ) . However , due to the limited number of appropriate relationships in GS10K to resolve correlations between matrices and to detect factors with small effects , our model selection procedure may omit minor effects ( contributing 5% or less of the trait variance , for example ) . In addition , the procedure may sometimes identify incorrect models ( not being able to distinguish familial effects as mentioned in the simulation study and S6 Table ) and this might be the case for weight , urea , TC and HR in Table 2 . However , with sufficient data from higher order pedigree relationships , as was the case in GS20K , the impact of covariances between design matrices in first order relatives ( parent-offspring , siblings and couples ) are mitigated and further components of variance became separable ( Table 3 ) . To sum up , we provide evidence that for the traits we have analysed , heritabilities are divided approximately evenly between pedigree-associated and SNP-associated genetic effects . This is the case even when , as here , we have taken care to consider various models of environmental covariation of first-degree relatives ( including couples ) . It appears that confounding factors like dominance , shared full-sibling environment and the past rearing environment seem to have relatively small contribution to phenotypic variation for these traits in our population . We find that current shared environment of couples is able to account for another ~11% on average of the phenotypic variation of human complex traits . This has been seldom mentioned in previous heritability studies but we note that as an effect that inflates the covariance between nominally unrelated individuals , it should not substantially bias or inflate hped2 and hgkin2 . It should be taken into account that couple effects may also be present in cohorts of unrelated individuals which may often include couples but ignore any correlation between them . Therefore , it might bias hg2 from genotype-based studies which do not account for such couple effects and could have an impact on GWAS studies . Overall , our work shows that SNP-associated genetic effects , pedigree-associated genetic effects and current shared couple environmental effects are three fundamental components of phenotypic variation for traits related to anthropometrics and cardiometabolism and current shared environmental effects have more impact than past shared environmental effects . This also has implications for models to be used in further studies of the architecture of complex traits including utilising the appropriate models for GWAS and related analyses and for personalised disease risk prediction .
The data were obtained from the Generation Scotland: Scottish Family Health Study ( GS:SFHS ) . Ethical approval for the study was given by the NHS Tayside committee on research ethics ( reference 05/s1401/89 ) and participants provided written consent . Governance of the study , including public engagement , protocol development and access arrangements , was overseen by an independent advisory board , established by the Scottish government Our dataset came from the Generation Scotland Scottish Family Health Study ( GS:SFHS ) project ( http://www . generationscotland . org ) , which was collected by a cross-disciplinary collaboration of Scottish medical schools and the National Health Service ( NHS ) from Feb 2006 to Mar 2011 [21 , 32] . Data for 16 complex traits were used . These were 8 anthropometric traits: height , weight , fat , body mass index ( BMI=WeightHeight2 ) , hips , waist , waist-to-hips ratio ( WHR ) and a body shape index ( ABSI =Waist Circumference×Height5/6Weight2/3 ) [20] and 8 cardiometabolic traits: levels of creatinine , urea , total cholesterol ( TC ) and high density lipoprotein ( HDL ) in serum and glucose in blood after a four hour fast period , systolic blood pressure ( SBP ) , diastolic blood pressure ( DBP ) and heart rate ( HR ) . None of the traits was adjusted for medication or fasting status . We explored the phenotypic distributions of these traits and conducted natural logarithm transformations for them , except for height , sodium and fat , to obtain approximate normal distributions . We set phenotypes with values greater or smaller than the mean ± 4 standard deviations ( after adjusting for sex , age and age2 ) to missing . Data also contained the information of sex , age , clinics where the phenotypes were measured and Scottish Index of Multiple Deprivation ( SIMD , an environmental ranking based on living areas , [33] ) . A descriptive analysis can be seen in S1 Table . The first set of analyses presented in the manuscript are based on a data set of nearly 10 , 000 individuals from GS:SFHS ( GS10K ) . These have multiple degrees of kinships , including 5 , 061 family members from 1 , 612 nuclear or extended families , and were genotyped with the Illumina OMNiExpress chip ( 707 , 686 SNPs ) . We conducted data quality control in Plink v1 . 07 [34] and GenABEL v1 . 7–6 [35] . SNPs with a minor allele frequency ( MAF ) < 0 . 05 , a Hardy-Weinberg Equilibrium’s ( HWE ) p-value <10−6 and a missingness > 2% were excluded . Duplicate samples , gender discrepancies and individuals with more than 5% missingness were also removed . After the quality control we kept 9 , 863 individuals genotyped for 550 , 796 common SNPs over the 22 autosomes . An extended dataset ( GS20K ) was used to validate the results obtained with GS10K and evaluate the effect of including further close relationships in our data . The extra 10 , 000 individuals were genotyped with the same chip and quality control was performed using the same criteria as in the GS10K . After quality control , GS20K consisted of 20 , 032 individuals , 18 , 293 of whom came from 6 , 578 nuclear or extended families , and 519 , 729 common SNPs across the 22 autosomes . A comparison of the difference in relationships between GS10K and GS20K can be seen in Table 1 . Our model allows trait variation to be influenced by the genetic effects associated with SNPs ( hg2 ) and pedigree ( hkin2 ) and the environmental effects shared by families ( ef2 ) , couples ( ec2 ) and full-siblings ( es2 ) , ( Fig 1 ) . To estimate the influence of each effect , we generated five design matrices: GRMg , GRMkin , ERMFamily , ERMSib and ERMCouple . A genomic relationship matrix ( GRM ) contains estimated genomic relatedness between pairs of individuals calculated from identity-by-state marker relationships as in Yang et al . [16 , 17] . Each off-diagonal entry in the GRM represents the realised genomic relationship between a pair of individuals: 1N∑i=1N ( xji−2pi ) ( xki−2pi ) 2pi ( 1−pi ) where , pi is the minor allele frequency ( MAF ) for SNP i , xji or xki is the allelic dose for individual j or k at locus i ( x = 2 if the individual carries two rare alleles , x = 1 if the individual is heterozygous , x = 0 if the individual carries two common alleles ) and N is the total number of SNPs . Each entry on the diagonal represents the inbreeding coefficient calculated as: 1+1N∑i=1Nxji2− ( 1+2pi ) xji+2pi22pi ( 1−pi ) We used GCTA [16] to generate GRMg and obtained GRMkin by modification of GRMg in R [36] . Their definitions are identical to matrices KIBS and KIBS>t in Zaitlen et al . [12] respectively . GRMg: a GRM estimated using all common SNPs , and designed to capture the additive genetic variance explained by common SNPs in the population sample . GRMkin: a modified GRM calculated as in Zaitlen et al . [12] designed to estimate the extra genetic effects associated with pedigree , the variance explained by shared genetic factors in close relatives . GRMkin was created by setting to 0 all entries in GRMg smaller than 0 . 025 . The number of entries different from 0 in each of the matrices is shown in Table 1 . An environmental relationship matrix ( ERM ) is a covariance matrix designed to capture the variance due to common environmental effects shared among a specified group of individuals . The ERM coefficient for each pair of individuals is 1 in if they share a particular environment , e . g . , living in the same area or coming from the same family; otherwise , it is 0 . Each entry on the diagonal is 1 . We generated 3 different ERMs in R [36]: ERMCouple , ERMSib and ERMFamily . ERMCouple: ERMCouple was designed to capture the common environmental effects shared between a couple . The ERM coefficient of two individuals was 1 if they were identified as a couple , defined as a pair of individuals with at least one offspring within GS:SFHS . Each entry on the diagonal was 1 . ERMSib: ERMSib was designed to capture the common environmental effects shared between full-siblings . The ERM coefficient of two individuals was 1 if they were identified as full-siblings . Each diagonal entry was 1 . ERMFamily: ERMFamily was designed to capture the common environmental effects shared within each nuclear family comprising parents and offspring . The ERM coefficient of two individuals was 1 if they were identified as a parent-offspring pair , full-siblings or a couple . The ERM coefficient of two individuals was 1 if they were identified as nuclear family members , including parent-offspring , couple and full-sibling relationships . Each diagonal entry was 1 . The number of entries different from 0 in each of the environmental matrices is shown in Table 1 . Details about model and matrices we defined can be seen in Fig 1 . We used the genomic and environmental matrices described above to partition the phenotypic variance observed for the traits using a mixed model in a restricted maximum likelihood ( REML ) framework . The analyses were implemented in GCTA [16] . The equations used to evaluate each model were the subsets of the full model: y=Xβ+gg+gkin+ef+es+ec+ε , with V=GRMgσg2+GRMkinσkin2+ERMFamilyσef2+ERMSibσes2+ERMCoupleσec2+Iσε2 where y is an n × 1 vector of observed phenotypes with n being the sample size ( number of individuals ) , and V the total phenotypic variance matrix , β is an m × 1 vector of fixed effects with m being the total level of covariates and X its design matrix with dimension n × m , gg is an n × 1 vector of the total additive genetic effects of the individuals captured by genotyped SNPs with gg∼N ( 0 , GRMgσg2 ) , gkin is an n × 1 vector of the extra genetic effects associated with the pedigree for relatives with gkin∼N ( 0 , GRMkinσkin2 ) , ef , es and ec are n × 1 vectors representing the common environmental effects shared by nuclear family members , full-siblings and couples with ef∼N ( 0 , ERMFamilyσef2 ) , es∼N ( 0 , ERMSibσes2 ) and ec∼N ( 0 , ERMCoupleσec2 ) and ε is an n × 1 vector of residuals . We fitted a range of models including different combinations of effects , and named them using abbreviations according to the effects used . We used the codes ‘G’ for GRMg , ‘K’ for GRMkin , ‘F’ for ERMFamily , ‘S’ for ERMSib and ‘C’ for ERMCouple –e . g . ‘GKC’ = GRMg + GRMkin + ERMCouple , and the proportion of total phenotypic variance captured by each matrix was termed hg2 , hkin2 , ef2 , es2 and ec2 accordingly . All models include a residual matrix and the total heritability hgkin2 is always the sum of hg2+hkin2 for any model . There were 31 different models from all the possible combinations of the five matrices . The abbreviations for each model and the formulae to estimate each term in each model are listed in S2 Table . The results for each model are listed in S4 Table . In addition to the matrices described ( including the residual matrix ) , we always included the fixed effects of sex , age , age2 , sex-by-age interaction , clinic , standardised SIMD and SIMD2 and the first 20 eigenvectors of GRMg ( to ameliorate problems associated with data structure ) . We conducted a stepwise model selection to find the most appropriate genetic and environmental model for each trait and dissect the phenotypic variation into its components ( SNP-associated additive genetic variance , pedigree-associated genetic effects shared among relatives and common environmental effects shared among the specified groups including nuclear family members , couples and full-siblings ) . The stepwise selection began with the full model ‘GKFSC’ , where all matrices were fitted together . We performed a Wald test and a log-likelihood ratio test ( LRT , using a mixture distribution of χdf=02 and χdf=12 with a probability of 0 . 5 [16] ) for each component and removed the component , if any , that was non-significant for both tests at α = 5% level and had the highest p-value for the Wald test . We repeated this process until all the remaining components were significant for at least one test . We did not correct for the limited number of traits analysed so error rates in this procedure should be considered to be on a per trait basis . In order to evaluate the robustness of our models and the performance of our stepwise model selection , we conducted a simulation study . We simulated , based on the real genotypic information and the real pedigree , different sets of phenotypes for each of the 9 , 863 individuals in GS10K . For simulating the genetic effects , we used a similar approach to Zaitlen et al . [12] by dividing the genome into two: even and odd chromosomes , and randomly selecting 550 SNPs from even and odd chromosomes ( approximately 1 from each 500 SNPs ) , representing the observed causal loci that were in LD with the SNPs ( SNP-associated genetic effects ) and the unobserved genetic variants that were not in LD with the SNP array ( pedigree-associated genetic effects ) separately . In a later step , only even chromosomes were used to generate GRMg and GRMkin . Each locus was assigned an effect size driven from exponential distribution as in Fisher [37] and the summed effects for even and odd chromosome SNPs were designed to explain hg2 and hkin2 of the trait variance respectively . For environmental factors , we simulated a sibling environmental effect , a couple environmental effect and two nuclear family environmental effects ( youth and adulthood environments ) for each individual . The corresponding effect sizes for sibling , couple and nuclear family environmental effects were derived from N ( 0 , es2 ) , N ( 0 , ec2 ) and N ( 0 , ef2 ) accordingly and were the same among full-siblings , between couples and among nuclear family members . In addition , we simulated a random residual effect for each individual , the residuals were derived from N ( 0 , ee2 ) where ee2 represents the proportion of variance remaining in each of the scenarios . For each scenario , each component ( hg2 , hkin2 , ec2 , es2 , ef2 ) was given a proportion of the variance explained and ee2 was 1−hg2−hkin2−ec2−es2−ef2 . The final phenotypes would be the sum of these genetic and environmental effects and residuals , and the expected mean and variance of simulated phenotypes were 0 and 1 , respectively . More details about how we simulated phenotypes can be found in S1 Text . We evaluated the robustness of our models under situations where phenotypes were contributed by i ) one of the five effects , ii ) SNP-associated genetic effects and one of the familial effects ( either pedigree-associated genetic effects or nuclear family environmental effects ) and iii ) SNP-associated genetic effects , familial effects and other environmental effects . All scenarios included residuals and 50 to 100 replicates were analysed for each scenario . The results of simulations were evaluated using a Z-test , which tested whether the mean estimate for each parameter deviated significantly from its simulated value . Note , it was too time consuming to explore all the possible combinations of models and simulated phenotypes , therefore , we mainly focused on the models that were selected in model selection procedure for the real phenotypes in GS10K ( Table 2 ) as well as the fundamental models of our study . More details about the parameter settings for these scenarios can be found in S5 Table . ERMFamily posited a relationship between siblings , parents-offspring and couples is somewhat confounded with the addition of GRMkin and ERMCouple , making separation and estimation of these effects ( ef2 , hkin2 and ec2 ) challenging , as confirmed by the results from analysis of real phenotypes in GS10K ( Table 2 ) . Hence , we evaluated the effectiveness of our model selection procedure under situations where phenotypes were contributed by moderate SNP-associated genetic effects and low sibling environmental effects plus a ) moderate nuclear family environmental effects but low pedigree-associated genetic effects and couple environmental effects , b ) low nuclear family environmental effects but moderate pedigree-associated genetic effects and couple environmental effects and c ) moderate nuclear family environmental effects , pedigree-associated genetic effects and couple environmental effects . All scenarios included residuals . More details about the parameter settings for these scenarios can be found in S6 Table . We conducted the model selection procedure for each replicate to see whether the final model selected matched the simulated phenotypic components for these scenarios ( Note: we ran 10 replicates for each scenario here ) . In addition , variance component analyses were performed using final selected models for these replicates to see whether the estimates of parameters were close to their simulated values . | Unravelling overall trait architecture of complex traits and diseases is important for phenotype prediction and disease prevention and correct modelling of the trait will further aid discovery of causative loci . Here we take advantage of genome-wide data and a large family-based study to examine the role of common genetic variants , pedigree-associated genetic variants , shared family environment , shared couple environment and shared sibling environment on 16 anthropometric and cardiometabolic traits . By analysing up to ~20 , 000 Scottish individuals , we find that common genetic variants , pedigree-associated genetic variants and recently-shared environment of couples are the most important contributors to variation in these traits , while past family and sibling environment have a limited impact . Further studies on the pedigree-associated genetic variation and the shared couple environment effect are needed , as little research has been devoted to them so far . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome-wide",
"association",
"studies",
"chemical",
"compounds",
"social",
"sciences",
"biomarkers",
"anthropology",
"organic",
"compounds",
"carbohydrates",
"glucose",
"simulation",
"and",
"modeling",
"genome",
"analysis",
"research",
"and",
"analysis",
"methods",
"anthropometry",
"chemistry",
"creatinine",
"biochemistry",
"physical",
"anthropology",
"organic",
"chemistry",
"phenotypes",
"heredity",
"genetics",
"monosaccharides",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"computational",
"biology",
"complex",
"traits",
"human",
"genetics"
] | 2016 | Pedigree- and SNP-Associated Genetics and Recent Environment are the Major Contributors to Anthropometric and Cardiometabolic Trait Variation |
Target-based screening is one of the major approaches in drug discovery . Besides the intended target , unexpected drug off-target interactions often occur , and many of them have not been recognized and characterized . The off-target interactions can be responsible for either therapeutic or side effects . Thus , identifying the genome-wide off-targets of lead compounds or existing drugs will be critical for designing effective and safe drugs , and providing new opportunities for drug repurposing . Although many computational methods have been developed to predict drug-target interactions , they are either less accurate than the one that we are proposing here or computationally too intensive , thereby limiting their capability for large-scale off-target identification . In addition , the performances of most machine learning based algorithms have been mainly evaluated to predict off-target interactions in the same gene family for hundreds of chemicals . It is not clear how these algorithms perform in terms of detecting off-targets across gene families on a proteome scale . Here , we are presenting a fast and accurate off-target prediction method , REMAP , which is based on a dual regularized one-class collaborative filtering algorithm , to explore continuous chemical space , protein space , and their interactome on a large scale . When tested in a reliable , extensive , and cross-gene family benchmark , REMAP outperforms the state-of-the-art methods . Furthermore , REMAP is highly scalable . It can screen a dataset of 200 thousands chemicals against 20 thousands proteins within 2 hours . Using the reconstructed genome-wide target profile as the fingerprint of a chemical compound , we predicted that seven FDA-approved drugs can be repurposed as novel anti-cancer therapies . The anti-cancer activity of six of them is supported by experimental evidences . Thus , REMAP is a valuable addition to the existing in silico toolbox for drug target identification , drug repurposing , phenotypic screening , and side effect prediction . The software and benchmark are available at https://github . com/hansaimlim/REMAP .
Conventional one-drug-one-gene drug discovery and drug development is a time-consuming and expensive process . It suffers from high attrition rate and possible unexpected post-market withdrawal [1] . It has been recognized that a drug rarely only binds to its intended target , and off-target interactions ( i . e . interactions between the drug and unintended targets ) are common [2] . The off-target interaction may lead to adverse drug reactions ( ADRs ) [3] , as demonstrated by the deadly side effect of a Fatty Acid Amide Hydrolase ( FAAH ) inhibitor in a recent clinical trial [4] . On the other hand , the off-target interaction may be therapeutically useful , thus providing opportunities for drug repurposing and polypharmacology [2] . Therefore , identifying off-target interactions is an important step in drug discovery and development in order to reduce the drug attrition rate and to accelerate the drug discovery and development process , and ultimately to make safer and more affordable drugs . Many efforts have been devoted to developing statistical machine learning methods for the prediction of unknown drug-target associations by screening large chemical and protein data sets [5] . One of the fundamental assumptions in applying statistical machine learning methods to drug-target interaction prediction is that similar chemicals bind to similar protein targets , and vice versa . Based on this similarity principle , both semi-supervised and supervised machine learning techniques have been applied . The semi-supervised learning methods either build statistical models for the k nearest neighbors ( k-NN ) of the query compound with similar compounds in the database ( e . g . Parzen-Rosenblatt Window ( PRW ) [6] and Set Ensemble Analysis ( SEA ) [7] are examples ) . Although a large number of 2D and 3D fingerprint representations of chemical structures have been developed , chemical structure similarity that is measured by Tanimoto coefficient ( TC ) or other similarity metrics of fingerprints is not continuously correlated with the binding activity . Activity cliff exists in the chemical space , where a small modification of a chemical structure can lead to a dramatic change in binding activity [8] . Thus , the chemical structural similarity alone is not sufficient to capture genome-wide target binding profile , as protein-chemical interaction is determined by both protein structures and chemical structures . New deep learning techniques that can learn non-linear , hierarchical relationships may provide new solutions for representing chemical space [9–12] . However , few work has been done to incorporate protein relationships into the deep learning framework . It remains to be seen whether the deep learning is applicable to genome-wide target prediction . A number of techniques such as Gaussian Interaction Profile ( GIP ) , Weighted Nearest Neighbor ( WNN ) , Regularized Least Squares ( RLS ) classifier [13 , 14] , and matrix factorization [15–17] have been developed to integrate chemical and genomic space . Among them , Neighborhood Regularized Logistic Matrix Factorization ( NRLMF ) [17] and Kernelized Bayesian Matrix Factorization ( KBMF ) [16] are two of the most successful methods . However , several drawbacks in these algorithms hinder their applications in genome-wide off-target predictions . First , several algorithms with high performance such as KBMF are extremely time and memory-consuming . Second , these algorithms depend on a supervised learning framework that requires negative cases . While publicly available biological and/or chemical databases ( e . g . ZINC [18] , ChEMBL [19] , DrugBank [20] , PubChem [21] , and UniProt [22] ) have enabled large-scale screening of drug-target associations , the known chemical-protein associations are sparse , and the number of reported negative cases ( i . e . chemical-protein pairs not associated ) is too small to optimally train a prediction algorithm [23] . Using randomly generated negative cases will adversely impact the performance of these algorithms , and algorithmically derived negative cases are often based on unrealistic assumptions [23] . Finally , these algorithms have been mainly evaluated for the prediction of off-targets within the same gene family ( e . g . GPCR ) using a small benchmark with hundreds of drugs and targets . Their performances in predicting off-target across gene families on a large scale are uncertain . Indeed , drug cross-reactivity often occurs across fold spaces [2] . Thus , the development of in silico prediction methods that are fast as well as accurate enough to explore the available data is urgent . Here , we make several contributions to address the aforementioned problems . First , we present an efficient method , REMAP , which formulates the off-target predictions as a dual-regularized One Class Collaborative Filtering ( OCCF ) problem . Thus , negative data are not needed for the training , but can be used if available . Secondly , REMAP is highly scalable with promising accuracy , thus can be applied to large-scale off-target predictions . Thirdly , we introduce a new benchmark set to evaluate the performance of drug-target interactions across gene families . Finally , we apply REMAP to repurposing existing drugs for new diseases . We identified seven drugs that have anti-cancer activity . Six of them are supported by experimental evidence .
The problem we try to solve here is to predict how likely it is that a chemical interacts with a target protein , using a chemical-protein association network , chemical-chemical similarity , and protein-protein similarity information . We start by preparing a bipartite network for chemical-protein associations as a sparse n × m matrix R , where n is the number of chemicals and m is the number of proteins . Ri , j = 1 if the ith chemical is associated with the jth protein , and Ri , j = 0 , otherwise . The chemical-chemical similarity scores are in an n × n square matrix C , with Ci , j representing the chemical-chemical similarity score between the ith and jth chemicals ( 0 ≤ Ci , j ≤ 1 ) for total n chemicals . The protein-protein similarity scores are in the same format for total m proteins ( 0 ≤ Ti , j ≤ 1 ) . We consider this problem an analog of user-item preferences such that users and items represent chemicals and proteins , respectively . Therefore , the problem is to provide an n × m matrix P in which Pi , j is the prediction score for the interaction between the ith chemical and the jth protein . Our prediction method REMAP is based on a one-class collaborative filtering algorithm that recommends the users’ preferences to the listed items [24] . It assumes that similar users will prefer similar items , unobserved associations are not necessarily negative , and user-item preferences can be analogous to drug-target associations . Assuming that a fairly low number of factors ( i . e . smaller number of features than the number of total chemicals or protein targets ) may capture the characteristics determining the chemical-protein associations , two low-rank matrices , U ( chemical side ) and V ( protein side ) , were approximated such that ∑in∑jm{R− ( U⋅VT ) } is minimized where R is the matrix for known chemical-protein associations and VT is the transposition of the protein side low-rank matrix V . The two low rank matrices , Un×r and Vm×r are obtained by iteratively minimizing the objective function , minU , V≥0∑ ( i , j ) pwt ( R ( i , j ) +pimp−U ( i , : ) ⋅V ( j , : ) T ) 2+preg ( ‖U‖2+‖V‖2 ) +pchemtr ( UT ( DC−C ) U ) +pprottr ( VT ( DT−T ) V ) ( 1 ) All symbols used in the paper are summarized in Table 1 , and the overall process of REMAP is in Fig 1 . Here , pwt is the penalty weight on the observed and unobserved associations which indicate the reliability of the assigned probability of true association , pimp is the imputed value ( i . e . the probability of unobserved associations as real associations ) , preg is the regularization parameter to prevent overfitting , pchem is the importance parameter for chemical-chemical similarity , pprot is the importance parameter for protein-protein similarity , and tr ( A ) is the trace of matrix A ( Table 1 ) . In this study , we use global weight and imputation . However , the weight and imputation values may be determined by a priori knowledge or from the prediction of other machine learning algorithms ( i . e . pwt and pimp can be matrices with the same dimension as the matrix R ) . The raw predicted score for the ith chemical to bind the jth protein can be calculated by P ( i , j ) =UUP ( i , : ) ⋅VUP ( j , : ) T . The raw scores were adjusted based on the ratio of observed positive and negative cases when the negative data are available ( explained in the prediction score adjustment section ) . Also , the matrix Un×r is referred to as a low-rank drug profile since its ith row represents the ith drug’s behavior in the drug-target interaction network as well as drug-drug similarity spaces compressed to r number of features . The REMAP code was originally written in Matlab and modified for drug-target predictions . Chemical-chemical similarity scores are one of the required inputs of REMAP . Although there are a number of metrics developed for chemical-chemical similarity , a recent study showed that Tanimoto coefficient-based similarity is highly efficient for fingerprint-based similarity measurement [25] . The fingerprint of choice in this study is the Extended Connectivity Fingerprint ( ECFP ) , which has been successfully applied to chemical structure-based target prediction method , PRW [6] . Thus , it allows for a fair comparison of REMAP with PRW . It is interesting to compare the different fingerprints in the future study . To calculate a similarity score between two chemicals , c1 and c2 , the Tanimoto dissimilarity coefficient dTani ( c1 , c2 ) was obtained using JChem with the Tanimoto metric for the ECFP descriptor type using the command in the Unix environment , “ChemAxon/JChem/bin/screenmd target_smi query_smi -k ECFP -g -c -M Tanimoto” [26] . The chemical-chemical similarity score , C ( c1 , c2 ) is defined as C ( c1 , c2 ) = 1-dTani ( c1 , c2 ) . Briefly , two chemicals have a higher similarity score if they have more of the same chemical moieties ( e . g . functional groups ) at more similar relative positions . Chemical similarity scores below 0 . 5 were treated as noise and set to 0 . Protein-protein similarity scores are also one of the required inputs for REMAP . The similarity between two proteins was calculated based on their sequence similarity using NCBI BLAST [27] with an e-value threshold of 1 × 10−5 and its default options ( e . g . 11 for gap open penalty and 1 for its extension , BLOSUM62 for the scoring matrix , and so on ) . Based on our 10-fold cross validation ( see below ) , e-value thresholds from 1 to 1 × 10−20 did not significantly affect the performance ( S1 Fig ) . Therefore , we decided to use a moderately stringent threshold ( BLAST default is 1 × 10−3 ) . A similarity score for query protein p1 to target protein p2 was calculated by the ratio of a bit score for the pair compared to the bit score of a self-query . To be specific , for the query protein p1 to the target protein p2 , protein-protein the similarity score was defined such that T ( p1 , p2 ) = dbit ( p1 , p2 ) /dbit ( p1 , p1 ) . For benchmark tests , ZINC data was filtered by IC50 ≤ 10 μM , which yielded 31 , 735 unique chemical-protein associations for 12 , 384 chemicals and 3 , 500 proteins ( ZINC dataset [18] ) . Targets that are protein complexes or cell-based tests were excluded . Proteins whose primary sequence is unavailable were also excluded . Protein sequences were obtained from UniProt [22] , and the whole protein sequences were used to calculate protein-protein similarity scores . To assess the predictive power of our algorithm , we performed a 10-fold cross validation on the ZINC dataset described above . We set the parameters as follows: pwt = pimp = preg = 0 . 1 , r = 300 , pchem = 0 . 75 , pprot = 0 . 1 , and piter = 400 . The optimized values determined by the 10-fold cross validation of benchmark are shown in S2 Fig . It is noted that the best performance is achieved when pchem = 0 . 25 and pprot = 0 . 25 . To further evaluate REMAP , we compared its performance on the ZINC dataset with several methods: a chemical similarity-based method ( PRW [6] ) , the best performed matrix factorization methods so far ( NRLMF [17] and KBMF with twin kernels ( KBMF2K ) [16] ) , combination of WNN and GIP ( WNNGIP [14] ) , and another type of matrix factorization method ( Collaborative Matrix Factorization ( CMF ) [15] ) for different types of chemicals and proteins . To obtain a detailed view of the performance of the methods , we divided the ZINC dataset into 3 categories with 2 subcategories for each , based on the connectivity of known chemical-protein associations and the degree of uniqueness of the chemicals . First , all the chemicals in the dataset were classified into the chemicals having only one known target ( NT1 ) , two known targets ( NT2 ) , or three or more known targets ( NT3 ) . Then , for the chemicals in each category , they were further divided based on either the number of known chemicals ( ligands ) the target proteins are associated with ( number of ligands in increments of 5 ) or the maximum chemical-chemical similarity score for the chemical in the dataset ( the similarity score range increment is 0 . 1 ) . The label used in this paper for the dataset are NTaLb , or NTaMaxTcd , where ‘NT’ stands for the Number of known Target , ‘L’ for the number of known Ligand , and ‘Tc’ for the maximum ( Tanimoto coefficient-based ) chemical-chemical similarity score for the given chemical in the dataset , with NT = a , b ≤ L ≤ b +4 , and d − 0 . 1 < Tc ≤ d . For instance , NT2L1 is the data set label for chemicals having two known targets and proteins having 1 to 5 ligands in the dataset , and NT1Tc0 . 9 is for chemicals with the most similar chemicals between 0 . 8 and 0 . 9 of similarity scores and having one known target . Chemicals having more than three known targets are included in the NT3 class , and proteins having more than twenty-one known ligands were included in L21 ( not limited to 25 ) . The categories of the ZINC dataset were then used to evaluate the performance of off-target prediction , and their labels mean the number of known ligands ( L ) or the maximum structural similarity ( Tc ) with their corresponding ranges . For example , ‘L21more’ stands for the dataset for proteins having 21 or more known targets , and ‘Tc0 . 9to1 . 0’ stands for maximum structural similarity greater than 0 . 9 and up to 1 . 0 ( Tc0 . 5to0 . 6 is inclusive of 0 . 5 ) . Note that NT1 is equivalent to chemicals without any known target when they are tested for cross validation . Therefore , performances on NT1 datasets reflect the ability to address the cold start problem . In other words , when one known drug-target association is intentionally hidden for the chemicals in the NT1 dataset , the tested chemicals will not have any known target in the training data , and they are less likely to be given a good recommendation of targets . This is analogous to the new user or new item problem reviewed by Su et al . [28] . A typical measure of prediction performance is the Receiver Operating Characteristic ( ROC ) curve by which one can assess the reliability of the positively predicted results . However , it is difficult to apply the ROC curve on our chemical-protein association datasets since the vast majority of the chemical-protein pairs have not been tested , and thus it is unclear whether the missing entries are actually unassociated or just not yet observed . In order to assess how reliable the positively predicted results from REMAP are , we needed to define a performance measurement that is analogous to ROC curve but not dependent on the true negatives . Our primary measure of performance is the true positive rate ( ∑True Positives∑Condition Positives; Recall or Recovery ) at the top 1% of predictions for each chemical . To be specific , the top 1% of predictions includes up to the 35th-ranked predicted target protein for a chemical for our datasets ( 3 , 500 possible target proteins for each chemical ) . Thus , for instance , a TPR of 0 . 965 at the 35th cutoff rank ( top 1% ) means that 96 . 5% of the total tested positive pairs were ranked 35th or better for the tested chemicals . In order to assess the speed of REMAP for practical uses , we measured its running time by varying the rank parameter or the size of dataset . On the ZINC dataset ( 12 , 384 chemicals and 3 , 500 proteins ) , up to r = 2 , 000 was tested , and at fixed r = 200 , dataset sizes up to 200 , 000 chemicals and 20 , 000 proteins were tested . The number of iterations ( piter ) was fixed to 400 . A single node of CPU with 2 . 88 GB of memory in the City University of New York High Performance Computing Center ( CUNY HPCC ) was used for REMAP running time tests . We also compared the running times of different matrix factorization methods with ours . Due to the large time complexity and memory requirement for other algorithms , a multi-core node with up to 700 GB of shared memory system in CUNY HPCC was used for them on the ZINC dataset . Chemical-protein associations were obtained from the ZINC [18] , ChEMBL [19] and DrugBank [20] databases . To obtain reliable chemical-protein association pairs , binding assays records with IC50 information were extracted from the databases , and the cutoff IC50 value of 10 μM was used where applicable . Two chemicals were considered the same if their InChI Keys are identical , and two proteins were considered so if their UniProt Accessions are identical . For records with IC50 in μg/L ( found in ChEMBL ) , the full molecular weights of the compounds listed on ChEMBL were used to convert μg/L to μM . Chemical-protein pairs were considered associated if IC50≤10 μM ( active pairs ) , unassociated if IC50>10 μM ( inactive pairs ) , ambiguous if records exist in both ranges ( ambiguous pairs ) , and unobserved otherwise ( unknown pairs ) . A total of 198 , 712 unique chemicals and 3 , 549 unique target proteins were obtained from the combination of ChEMBL and ZINC with 228 , 725 unique chemical-protein active pairs , 76 , 643 inactive pairs , and 4 , 068 ambiguous pairs . Of the 198 , 712 chemicals , 722 were found to be FDA-approved drugs . Furthermore , drug-target relationships were extracted from the DrugBank and integrated into the ZINC_ChEMBL dataset above . A total of 199 , 338 unique chemicals and 6 , 277 unique proteins were obtained from the combination of ZINC , ChEMBL , and DrugBank with 233 , 378 unique chemical-protein active pairs . Since REMAP showed promising performances on predicting off-targets for chemicals with at least one known target , it is possible to use REMAP to suggest new purposes for some FDA approved drugs . As the matrix product of UUP ( chemical-side low-rank matrix ) and VUP ( protein side low-rank matrix ) is the predicted drug-target interaction matrix P , the ith row of UUP contains the target interaction profile for the ith drug . Therefore , we analyzed the drug-drug similarities based on the low-rank matrix UUP . We ran REMAP with the data combination of three databases explained above , with the parameters used in the benchmark evaluations . Then , we calculated drug-drug cosine similarities based on the matrix UUP . For each row of UUP for FDA approved drugs , the cosine similarity of drug c1 and drug c2 can be calculated by , Scos , ( c1 , c2 ) = Uc1→∙Uc2→Uc1Uc2 . To search for possibly undiscovered uses of the drugs , we focused on drugs that are found to have high cosine similarity but low Tanimoto similarity ( < 0 . 5 ) . Markov Cluster ( MCL ) Algorithm [29 , 30] was used to cluster drugs based on their cosine similarity of a low-rank target profile . Drug-disease associations were obtained from the Comparative Toxicogenomics Database ( CTD ) [31] . The raw prediction score ( P ( i , j ) = UUP ( i , : ) ∙VUP ( j , : ) T ) can be adjusted to better reflect the real data as well as to statistically discriminate the positive and negative predictions . We used the active , inactive and ambiguous pairs obtained from the ChEMBL database to adjust the score . REMAP prediction on the ZINC_ChEMBL dataset showed a clear division between the active and inactive pairs , suggesting that predictions scored around 1 . 0 are highly likely to be positive ( Fig 2A ) . As mentioned above , however , there is a large difference between the number of active and inactive pairs , which is not likely to reflect the ratio of the actual positive and negative chemical-protein pairs . Greater accuracy is expected by adjusting the prediction scores to reflect such a positive/negative ratio . To estimate the ratio , we first normalized the counts in each bin in the histogram ( Fig 2A ) and calculated the weights that minimize the sum of error , Esum . Esum ( w1 ) = Σi[Ai − {w1pi + ( 1 − w1 ) Ni}]2 , where w1 and w2 are the weights on active and inactive pairs , respectively ( w1 + w2 = 1 . 0 ) , and Ai , pi and Ni are the normalized counts in ith bin of ambiguous , active and inactive pairs , respectively . The optimum adjustment weights were approximately w1 = 0 . 16 , w2 = 0 . 84 ( Fig 2B ) . This implies that approximately 16% of total observations are positive . Since the ratio of negative/positive is about 5 . 25 ( w2w1 = 5 . 25 ) , we increased the number of observations for inactive pairs in each bin by 5 . 25 times and rounded down . The adjusted prediction score for each bin ( Bi ) was calculated using the increased negative counts . It is noted that the prediction score adjustment was not used in the benchmark study , where no negative data were used . Drug-drug clustered network was visualized using Cytoscape [32] .
We evaluated the performances of algorithms for chemicals having one , two , or more than three known targets with varying maximum chemical-chemical similarity ranges or with proteins having a certain number of known ligands ( dataset prepared as explained in the methods and materials section ) . In general , the performances of both algorithms improve as the number of known ligands per protein or the maximum chemical-chemical similarity value increases . It was noticeable that REMAP performed significantly better than PRW when there was at least one known target for a chemical whose targets are predicted ( Figs 3 and 4 ) . REMAP showed greater than 90% recovery at the top 1% when the tested chemicals have at least one known target . All algorithms are sensitive to the number of ligands per target . The more ligands , the higher accuracy . While PRW also reached reasonably high recovery for some categories ( e . g . more than 11 known ligands per proteins , or C ( c1 , c2 ) >0 . 6 of the most similar trained chemicals ) , REMAP showed that it is reliable for testing chemicals without high similarity to the trained chemicals ( Figs 3B and 4B ) . In other words , REMAP is applicable to chemicals that are structurally distant to the chemicals already in the dataset . Except where the target proteins have 1 to 5 known ligands , REMAP performed best among the three algorithms in all cases with at least one known target for the tested chemicals ( Figs 3 and 4 ) . In the most of cases , the differences in the performance between REMAP and other two algorithms are statistically significant . Therefore , in practice , REMAP can predict potential drug targets for chemicals with at least one known target as training data , even when the chemicals are structurally dissimilar to the training chemicals . With the optimized parameters ( see below ) , ROC-like curves shows the general trend of performances of the three algorithms up to the top 10% of predictions ( S3 and S4 Figs ) . As shown in Figs 3 and 4 , REMAP outperforms the state-of-the-art NRLFM algorithm in most of the tested cases . As NRLMF is sensitive to the rank parameter , we carried out optimizations to determine optimal rank and iterations for NRLMF ( S5 Fig ) . The optimal rank and iterations used in the evaluation were 100 and 300 , respectively . Moreover , in the current implementation , REMAP is approximately 10 times faster and uses 50% less memory than NRLMF . Consistent with the results by Liu et al . [17] , the accuracies of NRLFM are significantly higher than KBMF2K , CMF , and WNNGIP in all of ZINC benchmarks . Overall , REMAP is one of the best-performing methods for the genome-wide off-target predictions . To test whether the chemical-chemical similarity matrix helps prediction , we performed 10-fold cross validation on the ZINC dataset with the contents of the chemical-chemical or the protein-protein similarity matrix controlled . In other words , about half of the non-zero chemical-chemical similarity scores were randomly chosen and removed ( set to 0 ) for the “half-filled chemical similarity” matrix , and all entries are set to 0 for the “zero-filled chemical similarity” matrix . The predictive power of REMAP showed noticeable improvement when all available chemical-chemical similarity pairs were used , compared to the half-filled or the zero-filled similarity matrix ( Fig 5A ) . Similarly , the contents of the protein-protein similarity matrix were controlled ( e . g . half-filled protein similarity , and zero-filled protein similarity ) while the full chemical similarity matrix was used . Unlike the chemical-chemical similarity , the protein-protein similarity information did not necessarily improve REMAP’s predictive power . The performance was best when a half of the protein-protein similarity information was used together with the full chemical-chemical similarity matrix ( Fig 5B ) . This suggests that there is significant noise in the protein-protein sequence similarity matrix although the information does help prediction . A careful examination of the BLAST-based protein-protein similarity matrix may give an insight into the design of a novel protein-protein similarity metric for drug-target binding activities ( see discussion section ) . We also performed optimization tests for pchem and pprot on ZINC dataset . Although the performance was slightly better when the chemical-chemical similarity importance was maximum ( Fig 6A ) , the difference was too small to conclude that it is best to fix pchem = 1 . Instead , the prediction may rely too much on the chemical-chemical similarity scores . Therefore , to allow flexibility on chemical-chemical similarity information , we set pchem = 0 . 75 at which the performance was almost as accurate as pchem = 1 . On the other hand , the performance was best when the protein-protein sequence similarity importance , pprot , was 0 . 1 ( Fig 6B ) , further supporting our claim that protein-protein sequence similarity is not an optimal choice for the prediction of a drug-target interaction . When jointly optimizing pchem and pprot , their optimal value is 0 . 25 and 0 . 25 , respectively , in the 10-fold cross validation benchmark evaluation ( S2B Fig ) . Our result supports a recent study [25] which showed that Tanimoto coefficient is efficient for the chemical similarity calculation . Chemical fingerprint-based chemical-protein association prediction has been studied by Koutsoukas et al [6] . By defining bins ( target proteins ) that can contain certain chemical features based on the chemical fingerprints , Koutsoukas et al . successfully demonstrated that their algorithm , PRW , can efficiently predict unknown chemical-protein associations [6] . While the basic idea of dissecting chemical compounds into functional groups is the same , it should be noted that PRW does not consider the information obtained from proteins , as well as interactome . For all our tests , REMAP showed great speed without losing its accuracy . On our benchmark dataset ( ZINC; 12 , 384 chemicals and 3 , 500 proteins ) , it took approximately 120 seconds to run 400 iterations at the rank of 200 ( r = 200 , piter = 400 ) . The time complexity is linearly dependent on the rank ( Fig 7A ) . The scalability of REMAP is superior when compared to KBMF2K , a state-of the art matrix factorization algorithm that is implemented in Matlab and has been extensively studied for predicting drug-target interactions [16] . KBMF2K took more than 10 days for the same size matrix using the same computer system in the ZINC benchmark . Moreover , REMAP was capable of higher rank factorization while KBMF2K was limited to rank 200 in our system due to the memory requirement ( over 100 GB of memory ) . At a much higher rank ( r = 2 , 000 ) , less than one hour was required for REMAP on the same dataset ( Fig 7A ) . Time complexity experiments on larger dataset showed that REMAP completed predictions on a dataset with 200 , 000 rows and 20 , 000 columns within 2 hours on a single core computing system with 2 . 88 GB of memory , demonstrating its ability to screen the whole human genome of approximately 20 , 000 proteins in two hours ( Fig 7B ) . Since REMAP is scalable and shows superior accuracy based on our benchmark tests , we performed large scale prediction of drug-target interactions on the ZCD dataset ( explained in the Materials and Methods section ) . As explained in the prediction score adjustment section , prediction scores for the active pairs were mostly located between 0 . 75 and 1 . 0 ( Fig 2A ) . As expected , the percentage of pairs of chemicals that share common targets decreases with the decrease of the chemical structural similarity measured by the Tc of ECFP fingerprints ( C ( c1 , c2 ) ) . The percentage of target-sharing chemical pairs drops below 50% and 0 . 5% when the Tc is between 0 . 5 and 0 . 6 , and less than 0 . 5 , respectively ( S6 Fig ) . Thus , it is less likely that the chemical structural similarity alone can reliably detect novel binding relations between two chemicals when the Tc is less than 0 . 5 . It is interesting to see how REMAP performs when the chemical structural similarity fails . We analyzed the low-rank drug profile ( matrix UUP ) to check whether it represented the target-binding behavior of the drugs . When filtered by low chemical structure similarity ( C ( c1 , c2 ) <0 . 5 ) ) , there are 899 , 871 drug-drug pairs . Among them , the profile similarity score ( Scos , ( c1 , c2 ) ) of 91 , 888 pairs is higher than 0 . 3 . With high profile similarity ( 0 . 99≤Scos , ( c1 , c2 ) ≤1 ) ) , a total of 1 , 327 drug-drug pairs were found of which 1 , 033 pairs shared at least one common known target . S7 Fig shows the percentage of pairs that share the common target in different profile similarity bucket for FDA-approved drugs . This result suggests that REMAP is able to provide a chemical-protein binding profile that cannot be captured by chemical structure similarity alone . When Scos , ( c1 , c2 ) ≤0 . 3 , the percentage of two drugs that share a common target drops below 50% ( S7 Fig ) . We constructed a drug-drug similarity network by filtering out drug pairs with Scos , ( c1 , c2 ) ≤0 . 3 , then applied the MCL algorithm on the drug-drug network to find clusters of similar drugs . The largest cluster of drugs contained a total of 313 drugs , and their relationships to diseases were examined based on the known associations annotated in CTD [31] . As a result , we found that the drugs are mostly related to mental disorders , including hyperkinesis , dystonia , catalepsy , schizophrenia and basal ganglia diseases as the mostly related diseases . The most frequent known protein targets by the drugs were GPCRs ( S1 Table ) . It is comparable that GPCRs were 1 , 924 times targeted while kinases were targeted only 55 times . While it is interesting to further examine the cluster , validating all of the possible drug-target pairs in the largest cluster may be inefficient . A smaller cluster of drugs contained a total of thirty-one FDA approved drugs twenty-six of which are known to target kinases or interact with microtubule ( Table 2 ) . Seven drugs in the cluster have not been used for cancer treatment and were found to be closely linked to the anti-cancer drugs ( Fig 8 and Table 2 ) . Interestingly , several of them have been tested for their anti-cancer activity . For example , colchicine ( also known as colchine ) , an FDA approved drug for gout treatment , has been shown to have anti-proliferative effects on several human liver cancer cell lines at clinically acceptable concentrations [33] . Griseofulvin , an antifungal antibiotic drug , appears to be effective as an anti-cancer drug when used together with other anti-cancer drugs [34] . The three anthelmintic drugs , albendazole , mebendazole and niclosamide , have been studied and repurposed for their anti-cancer effects on different types of cancers . Albendazole has been shown to be effective in suppressing liver cancer cells both in vitro and in vivo [35] , and recently has been repurposed for ovarian cancer treatment with a bovine serum albumin-based nanoparticle drug delivery system [36] . Mebendazole showed anti-cancer activities in human lung cancer cell lines [37] and human adrenocortical cell lines [38] , and it has been repurposed for colon cancer treatment [39] . Both niclosamide and mebendazole showed beneficial effects in glioblastoma in different studies [40 , 41] . It has been proposed to use aprepitant in combination with other compounds to improve the efficiency of temozolomide , the current standard drug for glioblastoma treatment [42] . Anti-cancer activity of carbidopa hydrate have not yet been reported . It will be interesting to experimentally validate the prediction .
Our extensive benchmark studies show that REMAP outperforms existing algorithms in most of the cases for the off-target prediction . Compared with other state-of-the-art matrix factorization algorithms , the predictive power of REMAP comes from several improvements . First , we formulated the drug-target prediction as a one-class collaborative filtering problem; thus the negative data are not required for the training . Second , a priori knowledge including known negative data can be incorporated into the matrix factorization with imputation and weighting . Finally , using global imputation and weighting , the algorithm is computationally efficient without significantly sacrificing its performance . The efficiency and effectiveness of REMAP allows us to predict proteome-wide target binding profiles of hundreds of thousands of chemicals . As the proteome-wide target binding profile is more correlated with phenotypic response than a single target binding , REMAP will facilitate linking molecular interactions in the test tube with in vivo drug activity . When using a multi-target binding profile predicted by REMAP as the signature of a chemical compound , seven drugs were found to be associated with anti-cancer therapeutics , although they do not have detectable chemical structural similarity . Among them , the anti-cancer activity of six drugs was supported by experimental evidences . Thus , REMAP could be a useful tool for drug repurposing . Although REMAP showed its high potential on genome-wide off-target predictions as discussed above , two issues remain: the cold start problem and suboptimal protein-protein similarity metrics . Similar to matrix factorization algorithms such as NRLMF , REMAP suffers from cold start problem , also known as new user or new item problem . In other words , it is difficult to recommend a product for a new user if the new user has never purchased or reviewed a product in the database [28] . For novel chemicals that do not have any known target in the dataset , REMAP did not show better performance than PRW . Moreover , if the target of the novel chemical has 5 or fewer known ligands , the recovery of REMAP is lower than 0 . 5 ( S8A Fig ) . When the novel chemical is similar to those chemicals in the database , the recovery of REMAP reached above 90% ( S8B Fig ) . These results suggest that , in practice , existing matrix factorization-based methods , including REMAP , are not the optimal choice if the chemicals of interest do not have any known target . To resolve this issue , it is possible to design an algorithm that combines the benefits of PRW or other algorithms with REMAP . The use of confidence weights and a priori imputation makes it straightforward for REMAP to incorporate additional information . In addition , the time and memory efficiency of REMAP makes it possible to apply active learning to overcome the cold start problem [43–46] . The suboptimal performance of REMAP may arise from the lack of molecular-level biochemical details in deriving the protein-protein similarity metrics . When testing the ZINC dataset , we found that REMAP performs better as lower weight was assigned for protein-protein sequence similarity data ( Fig 6B ) . In addition , the predictive power of REMAP improved when about half of the randomly selected protein-protein similarity scores were removed , further confirming that noise confounds relating global sequence similarity to ligand binding ( Fig 5B ) . It is not surprising that proteins with similar sequences do not necessarily bind to similar chemicals , as protein-ligand interaction is governed by the spatial organization of amino acid residues in the protein structure [47] . Amino acid mutations/post-translational modifications and conformational dynamics may alter the binding of the ligand through direct modification of the ligand binding site or allosteric interaction . A protein may also consist of multiple binding sites that accommodate different types of ligands . Thus , two proteins with high sequence similarity do not necessarily bind the same ligands because the two proteins may possess different 3D conformations , especially in their binding pockets [47] . In contrast , two proteins with low sequence similarity can bind to the same ligands if their binding pockets are similar [48 , 49] . The binding site similarity can be a more biologically sensitive measure of protein-protein similarity for the off-target prediction [50–55] . Such work is on-going . In silico drug-target screening is an essential step to reduce costly experimental steps in drug development . In this study , we showed that dual-regularized one-class collaborative filtering algorithm , a class of computational methods frequently used in user-item preference recommendations , may be applied to drug-target association predictions . Our study presents REMAP , a collaborative filtering algorithm with capability of running whole human genome-level predictions within two hours . Other studies on some types of cancer treatment support our algorithm’s ability to capture drug-drug similarities based on both the drug-target interaction profile and the chemical structural similarity . Our study shows the limitation of REMAP in evaluating new chemicals or accommodating biochemical details . Further development of the computational tools for better prediction is needed . | High-throughput techniques have generated vast amounts of diverse omics and phenotypic data . However , these sets of data have not yet been fully explored to improve the effectiveness and efficiency of drug discovery , a process which has traditionally adopted a one-drug-one-gene paradigm . Consequently , the cost of bringing a drug to market is astounding and the failure rate is daunting . The failure of the target-based drug discovery is in large part due to the fact that a drug rarely interacts only with its intended receptor , but also generally binds to other receptors . To rationally design potent and safe therapeutics , we need to identify all the possible cellular proteins interacting with a drug in an organism . Existing experimental techniques are not sufficient to address this problem , and will benefit from computational modeling . However , it is a daunting task to reliably screen millions of chemicals against hundreds of thousands of proteins . Here , we introduce a fast and accurate method REMAP for large-scale predictions of drug-target interactions . REMAP outperforms state-of-the-art algorithms in terms of both speed and accuracy , and has been successfully applied to drug repurposing . Thus , REMAP may have broad applications in drug discovery . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"applied",
"mathematics",
"cancer",
"treatment",
"enzymology",
"simulation",
"and",
"modeling",
"oncology",
"algorithms",
"genomic",
"databases",
"tyrosine",
"kinase",
"inhibitors",
"mathematics",
"forecasting",
"statistics",
"(mathematics)",
"genome",
"analysis",
"pharmacology",
"enzyme",
"inhibitors",
"research",
"and",
"analysis",
"methods",
"genomics",
"mathematical",
"and",
"statistical",
"techniques",
"biological",
"databases",
"drug",
"discovery",
"biochemistry",
"kinase",
"inhibitors",
"drug-drug",
"interactions",
"drug",
"research",
"and",
"development",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"drug",
"interactions",
"statistical",
"methods"
] | 2016 | Large-Scale Off-Target Identification Using Fast and Accurate Dual Regularized One-Class Collaborative Filtering and Its Application to Drug Repurposing |
Sensory information about the state of the world is generally ambiguous . Understanding how the nervous system resolves such ambiguities to infer the actual state of the world is a central quest for sensory neuroscience . However , the computational principles of perceptual disambiguation are still poorly understood: What drives perceptual decision-making between multiple equally valid solutions ? Here we investigate how humans gather and combine sensory information–within and across modalities–to disambiguate motion perception in an ambiguous audiovisual display , where two moving stimuli could appear as either streaming through , or bouncing off each other . By combining psychophysical classification tasks with reverse correlation analyses , we identified the particular spatiotemporal stimulus patterns that elicit a stream or a bounce percept , respectively . From that , we developed and tested a computational model for uni- and multi-sensory perceptual disambiguation that tightly replicates human performance . Specifically , disambiguation relies on knowledge of prototypical bouncing events that contain characteristic patterns of motion energy in the dynamic visual display . Next , the visual information is linearly integrated with auditory cues and prior knowledge about the history of recent perceptual interpretations . What is more , we demonstrate that perceptual decision-making with ambiguous displays is systematically driven by noise , whose random patterns not only promote alternation , but also provide signal-like information that biases perception in highly predictable fashion .
Perception is well described as an inference process based on noisy and often ambiguous sensory signals . As such , a single sensory modality most often cannot provide enough information to univocally specify the actual state of the world . Throughout the history of vision science , numerous ambiguous displays have been put forward where the very same sensory stimulus allows for multiple , and clearly distinguishable , alternative interpretations—multistable stimuli such as the Necker Cube , the stream-bounce display and binocular rivalry [1–4] . However , in our daily lives perception seems to be surprisingly devoid of such ambiguities . This is because in most real-life scenarios , the brain can rely on a large variety of information that is often not present in the minimalistic ambiguous displays used in laboratory settings . Information for perceptual disambiguation can come from different cues derived from the same or other senses [5] , or it may come in form of prior knowledge representing the statistical regularities found in the natural world [6 , 7] , or recent perceptual history [8–15] . A notable example of ambiguous stimulus is the stream-bounce display ( cf . S1 Movie ) : two identical objects moving towards each other along the same trajectory can be perceived as streaming through each other , or as colliding and bouncing away from one another [1] . Vision alone does not provide enough information to rule out any possible interpretation , and over repeated presentations the two percepts alternate in a seemingly arbitrary fashion . However , if a sound is presented around the time of the crossing , humans are more likely to perceive a bounce [5] . This finding demonstrates that humans integrate multisensory information for perceptual disambiguation , that is , to infer the most likely interpretation of the sensory data . However , the underlying mechanism of this inference process is still poorly understood . What drives perceptual decision making , and which strategy does the brain use to extract and combine information from the different senses ? Sensory information is corrupted by noise arising in the brain at any stage of processing [16 , 17] . Therefore , a possible reason for perceptual alternation ( in cases where the two states are about equally likely ) relies on the random fluctuations of the internal noise [18] . The role of noise on ambiguous displays has been widely investigated over the years; for example , motion coherence ( i . e . , noise in the motion signal ) can reliably predict the time-course of perceptual switches in binocular rivalry with moving stimuli [19] . Likewise , internal noise is at the heart of most computational models for perceptual alternation in binocular rivalry [19–24] . Specifically , perceptual alternation is often interpreted in terms of a double-well energy landscape [24] , where both adaptation-recovery and noise contribute perceptual shifts [21] . Moreover , the effects of noise on binocular rivalry have been successfully simulated with biologically plausible spiking neuron models [20 , 23] . While these models can predict perception in binocular rivalry given the statistical properties of the noise , it is currently unclear how the individual instances of the noise affect perceptual disambiguation . For example , the spatiotemporal patterns present in the noise may contain signal-like information that systematically biases perception towards one specific interpretation ( and against its alternative ) . Unfortunately , scientists do not have direct access to the noise that is present within the brain , thus making it hard to test this hypothesis . To overcome this problem , and systematically study the effects of noise on perception , psychophysicists often try to override the noise in the system by introducing ( external ) noise directly in the stimulus [25] . Given that noise does not come with a label , the brain often cannot infer its internal vs . external nature , so it is reasonable to assume that the brain usually treats these two sources of noise in the same way [17] , ( though see [26] for a recent finding on how the brain may sometimes be capable of distinguishing , and filtering out , different types of noise ) . Once the exact properties of the noise are known , reverse correlation techniques [27–29] can be used to investigate whether the brain looks for certain patterns in the noise that might help resolving ambiguity . That is , by investigating how the distribution of external noise on each given trial biases perception , it becomes possible–by averaging the classified stimuli and their noises–to estimate the spatiotemporal decision template used for perceptual disambiguation . In turn , this allows one to explore a number of important aspects of perceptual disambiguation , such as what biases perception , and how sensory information is combined across the senses–and over time and space–to determine the final percept . In the present study we used reverse correlation techniques to investigate the multisensory mechanisms for perceptual disambiguation in the stream-bounce display . Research on multisensory integration has demonstrated that when integrating redundant and unambiguous information from different sensory modalities , the brain operates in a statistically optimal fashion by taking a weighted average of the individual sensory cues . Thereby , the weights are assigned according to the precision of each cue [30] . This solution is statistically optimal because it provides the most accurate and precise perceptual estimate given the noisy sensory information as input . In the case of the stream-bounce display , however , the information provided by vision and audition is complementary in nature . More specifically , while vision informs us about the spatiotemporal trajectories of the moving objects ( while information about the nature of the impact is ambiguous: present or not ) , audition informs us about the presence of an impact by an appropriately timed sound ( or the absence of an impact if the sound is absent or presented with inappropriate timing ) . That is , vision and audition provide information in qualitatively different formats , which cannot be directly averaged without further transformations ( e . g . , see [31] ) . The aim of the present study is to characterize how the brain extracts , transforms , and combines complementary information within and across the senses to resolve perceptual ambiguity .
Three participants ( the author CP , age 33 , male; and two female naïve observers , CG and VL age 24 and 23 , respectively ) were presented with two small vertical light gray bars ( 0 . 085° x 0 . 426° each ) moving horizontally along the same trajectory in opposite directions , crossing at the center ( signal: Fig 1 , left ) . The moving stimuli were embedded in dynamic visual noise ( noise: Fig 1 , center; signal+noise: Fig 1 , right; S2 Movie ) randomly increasing or decreasing in luminance from the middle grey background . In a forced choice task , participant had to report whether they perceived the bars as streaming across each other or as colliding and bouncing off each other ( see Methods ) . The experiment consisted of ~10 , 000 trials per participant with the dynamic visual noise randomly varying across trials . In half of the trials , a sound ( 10ms white noise click ) was presented at the time of the crossing . Sound and no-sound trials alternated throughout the experiment in blocks of 40 trials . Overall , participants perceived the stimulus as bouncing on 43% of the trial ( CP: 48%; CG: 40%; VL: 42% ) . In line with previous studies , sounds significantly modulated participants’ responses and systematically biased the percept toward a bounce: only 27% of the trials without sound were perceived as bouncing , against 61% of the trials with sounds ( Fig 2C ) . Also , participants had a strong tendency towards interpreting the stimulus just as they did in the previous trial ( Fig 2C ) . Such a perceptual stability over time is a classic finding in the study of ambiguous displays [13 , 15 , 21 , 32] , and it has recently been demonstrated also for the stream-bounce display [33] . This shows that perceptual disambiguation also relies on the combination of current sensory information with memory of recent perceptual interpretations [34 , 35] . However , it should be noted that this effect might be partially due to the fact that sound and no sound trials were presented in separate blocks , thus stabilizing the percept within each block . The responses of motion energy filters are strongly modulated by the contrast of the moving object , while being relatively insensitive to luminance . Therefore , a critical test for the role of motion energy filters in disambiguating the stream-bounce display would be to invert the luminance polarity of the moving bars with respect to the background , while keeping their contrast constant . If motion energy computation is indeed the underlying mechanism for perceptual disambiguation , it should not matter whether the moving bars are lighter than the background ( like in the previous experiment ) or darker . Rather , what would matter should be the amount of the drop of motion energy at the intersection , which correlates with the total motion energy in the display . To directly test this hypothesis , we generated stimuli analogous to the ones used in the previous experiment , but with an additional modulation of both the total motion energy and the luminance polarity of the moving bars ( Fig 5A and S3 Movie , see Methods for further details ) . Next , we run a psychophysical task where we asked participants to classify such displays as streaming or bouncing , with the hypothesis that stimuli with high motion energy , and hence with a large drop of motion energy at the intersection , should be more likely classified as bouncing , irrespective of the luminance polarity of the moving bars . As hypothesized , displays with a large drop in motion energy ( i . e . , high motion energy ) were systematically classified as bouncing more often than those with a smaller drop ( i . e . , low motion energy , Fig 5B ) , whereas luminance did not significantly affect participants’ responses . Notably , the magnitude of the effect of motion energy drop on perceptual disambiguation was comparable to the effect of sound presence/absence in Experiment 1 ( compare Fig 2C left , and Fig 5B ) , and it was highly consistent across participants ( Fig 5C ) . This result demonstrates that motion energy , and its drop at the intersection , is indeed the key visual factor driving perceptual disambiguation , and further validates the current classification model .
The mechanisms underlying perceptual disambiguation are a central topic in sensory neuroscience . Resolving ambiguity requires both extracting and combining sensory information . The present results demonstrate which cues are relevant to resolve perceptual ambiguity in the stream-bounce display , and highlight the mechanisms underlying both the computation and the combination of such multisensory cues for the perception of dynamic ambiguous displays . Previous research has already investigated the stimulus properties biasing the interpretation of the stream-bounce display ( e . g . , [36 , 38 , 46] ) . Such studies relied on a parametric manipulation of one or more stimulus features . However , this approach requires researchers to decide a priori which features might be relevant to solve the task . The key advantage of using noisy stimuli and reverse correlation analyses relies on the absence of any prior assumptions , which allows us to determine a-posteriori how random fluctuations in the ( external ) noise systematically modulate participants’ responses . This , in turn , allows gathering detailed information about the precise spatiotemporal features buried in the stimuli that underlie perceptual disambiguation . Such key features and their relative contribution to perceptual decision making can only be obtained using standard psychophysical procedures by lucky guessing , and it is never clear whether some key features have been missed . An example from this study is the importance for disambiguation of motion energy , and its drop at the intersection , which might not have been discovered with traditional psychophysical methods . What is more , reverse correlation analyses revealed that sound does not alter early visual processing; rather , it modulates the percept after the unimodal information for disambiguation has been independently extracted from all modalities . Over the last decade , multisensory integration has often been modeled in terms Bayesian decision theory . Empirical results demonstrate that the brain operates in a statistically optimal way by maximizing the accuracy and precision of combined sensory estimates when integrating redundant and unambiguous sensory information [30] . Before integrating multisensory information for perceptual disambiguation , however , the brain needs to transform sensory information into a common format to make it directly comparable . That is , the different signals should be separately processed to extract stream-bounce information ( i . e . , probability of impact ) from the continuous stream of sensory signals . Here , we modeled this transformation in terms of motion energy filters [40] , which transforms complex , dynamic visual information into proxy decision variables that represent the strength of sensory evidence . Once transformed into a common format , sensory evidence from vision and audition can be directly compared and integrated by weighted averaging . A simple linear integration model ( without interactions across the cues ) captures human perception with a high degree of accuracy [36] . This result demonstrates that the brain applies analogous integration principles for disambiguation as it does for integrating redundant information [44] . What is more , the close correspondence between the empirical classification images and those predicted based on the motion energy model , further supports the motion energy model itself . A pressing question in the study of perceptual ambiguities concerns the conditions or parameters that drive perceptual biases . That is , why on each trial a given interpretation is chosen over the competing one . Internal noise has often been advocated to explain perceptual alternation [19 , 21–24] . Namely , noise was advocated as causing perceptual switches during prolonged presentations of bistable stimuli , like in binocular rivalry . However , to date we still do not know exactly which spatiotemporal ( i . e . , “signal-like” ) patterns within the noise selectively drive perceptual disambiguation between two equally valid alternatives . By embedding the stimulus in external noise with known properties , and using reverse correlation analyses , this study demonstrates not only that noise is indeed the key element driving alternation , but also which pixel-by-pixel properties of the noise are systematically associated to each interpretation . More specifically , due to its random structure , noise often contains information that is used in the process of resolving ambiguity . Here , we characterize what such properties are in the case of the stream-bounce display , and how they get extracted though spatiotemporal visual filters . Notably , the existence of systematic links between low-level stimulus properties and perceptual responses–as measured through reverse correlation analyses–argues against interpretations of disambiguation purely in terms of attention or response biases [47–50] . In recent years , the stream-bounce display has often been used to investigate the neural underpinnings of multisensory integration . The main findings support the involvement of multimodal cortical regions [51] and large-scale synchronizations of oscillatory neural activity [52 , 53] in resolving multisensory perceptual ambiguity . However , the computational principles underlying the multi-stability of the stream-bounce display remained poorly understood . The current results fill this important gap and provide evidence for the fundamental role of motion energy computation and linear integration of evidence in multisensory perceptual disambiguation .
Three participants ( 2 naïve females , VL and CG , and one male , the author CP ) took part in Experiment 1 . All participants were right handed and had normal or corrected to normal vision and audition . Participants sat in front of a computer screen with their head constrained by a chin and headrest . Each trial started with the presentation of a red fixation cross at the center of the screen ( 600ms ) , after which the visual stimulus appeared , consisting of two light vertical bars ( 0 . 085° x 0 . 426° each ) moving in opposite direction and embedded in dynamic visual noise . The dynamic stimulus comprised 20 frames ( 60Hz screen , overall duration 333ms ) of uni-dimensional visual noise consisting of 20 vertical bars ( 0 . 085° x 0 . 426° each ) with random luminance . The luminance of each noise sample varied between 14 . 6 cd/m2 and 48 . 3 cd/m2 . The moving visual stimuli were defined by a 50 cd/m2 luminance increment . The stimuli used in Experiment 1 are contained in S1 Dataset . On half of the blocks , a 16ms white noise auditory click was played from two speakers flanking the screen when the two moving stimuli met at the center of the screen . Participants were informed about the presence or absence of the sound at the beginning of each block . Such a blocked design was necessary as in preliminary observations we found that when sound and no-sound trials alternated randomly , participants’ responses were almost-exclusively determined by the presence/absence of the sound . Therefore , a blocked design made it easier to empirically estimate visual classification images . The relatively small size and short duration of the stimuli were selected to discourage eye-movements . Observers’ task was to look at the stimuli without moving their eyes , and to report by a key-press whether they perceived the stimuli as bouncing or streaming through each other . Participants were explicitly told that there was no correct or wrong answer . The experiment was performed in a dark anechoic chamber and it was controlled by a custom-built software based on the Psychtoolbox 3 [54] . Experiment 1 consisted of ~10 , 000 trials per participant ( CP: 10240 trials; CG: 10320; VL: 9840 trials ) . Psychophysical data is contained in S2 Dataset . Sound significantly modulated the percept ( overall: χ2 = 3404; p<0 . 001; CP: χ2 = 584; p<0 . 001; CG: χ2 = 1503; p<0 . 001; VL: χ2 = 1491; p<0 . 001 ) : only 27% ( CP: 36%; CG: 21%; VL: 23% ) of the trials without sound were perceived as bouncing , against 61% ( CP: 60%; CG: 59%; VL: 62% ) of the trials with sounds ( Fig 2C ) . Also , participants had a strong bias towards interpreting the stimulus just as they did in the previous trial ( overall χ2 = 2693; p<0 . 001; CP: χ2 = 531; p<0 . 001; CG: χ2 = 1818; p<0 . 001; VL: χ2 = 547; p<0 . 001 ) . Given that we were interested on the effects of the previous response , the first trial of each block ( of 40 trials ) was discarded from these analyses . In Experiment 2 we used the motion energy model to generate noisy stream-bounce displays–with dark and light moving bars–with a parametric manipulation of motion energy ( and hence of motion energy drop , see S3 Movie ) . This was achieved by first calculating classification images based on total motion energy for noisy visual stimuli with both light and dark moving bars . To estimate the classification images for motion energy , we generated a series of 400 , 000 noisy stimuli like in the previous experiment , with the only difference that the moving bars could be either lighter or darker than the background . Such stimuli were then fed into the motion energy model ( see below ) to calculate their total motion energy ( i . e . , the sum of rightward and leftward motion energy ) , and finally we discretized the models’ response by classifying the 50% of the stimuli with higher motion energy as bounce and the remaining ones as streaming . This procedure was separately performed for stimuli with light and dark bars . Classification images were then calculated following the same procedure used in Experiment 1 . The resulting classification images looked similar to those of Experiment 1 ( Fig 4A ) , however , the luminance kernels for light and dark bars had opposite polarities . Such classification images were then used to create stream-bounce displays with high or low motion energy ( S3 Movie ) . First off , we generated a series of stimuli similar to the ones used in Experiment 1 , but again the moving bars could be either lighter or darker than the background . Then , we experimentally manipulated the amount of motion energy by either adding or subtracting the luminance kernels to obtain displays with high or low motion energy , respectively . This procedure was separately performed for light and dark bars using their respective classification images . Such manipulated displays were then used in a psychophysical classification task in Experiment 2 . Overall , the task was very similar to Experiment 1 . However , in Experiment 2 we did not play any sounds . Visual stimuli with high and low motion energy randomly alternated across trials , while light and dark bars were presented in separate blocks of 16 trials each . In total , the experiment consisted of 126 trials . That is , 32 trials for each of the four combinations of bars’ luminance ( light vs . dark ) and total motion energy ( high vs . low ) . Ten naïve participants ( 7 females ) took part in Experiment 2 . Compared to Experiment 1 , the larger number of participants in Experiment 2 was due to the fact that in the latter experiment each participant performed a much smaller number of trials ( 126 vs . ~10 , 000 trials ) . Before starting the experiment , all participants underwent a short practice session to familiarize with the stimuli and the task . The probability of reporting a bounce for each condition and participant was normalized through a Z-score transformation and was analyzed using a 2x2 repeated-measures ANOVA with motion energy and luminance as within-participants factors . Motion energy strongly modulated participants’ responses ( F ( 1 , 9 ) = 35 . 489 , p<0 . 001 ) , with no effects of luminance ( F ( 1 , 9 ) = 2 . 317 , p = 0 . 162 ) and no interactions ( F ( 1 , 9 ) = 1 . 036 , p = 0 . 335 ) . This study was conducted in accordance with the Declaration of Helsinki and the experiments had ethical approval from the ethics committee of the University of Tübingen . To calculate visual classification images we first sorted the noisy visual stimuli presented in the experiment according to participants’ ( or model’s ) classification responses ( stream vs . bounce ) . For each class , we calculated the mean luminance ( μ ) and contrast ( mean squared error , MSE ) of the noisy visual stimuli and we combined them as follows to obtain the classification images for visual luminance ( KL ) and contrast ( KC ) : KL ( x , t ) =μ[bounce] ( x , t ) −μ[stream] ( x , t ) ( 1 ) KC ( x , t ) =MSE[bounce] ( x , t ) −MSE[stream] ( x , t ) ( 2 ) Where μ[R] and MSE[R] are the mean and the mean squared error templates for the stimuli S[R] , respectively . R denotes participants’ responses . Visual classification images ( KL , KC ) were smoothed by convolution with a low-pass spatiotemporal filter of the form [0 . 49 , 0 . 7 , 0 . 49; 0 . 70 , 1 . 0 , 0 . 70; 0 . 49 , 0 . 7 , 0 . 49] [55] . Finally , all classification images were range-scaled so that their maximum absolute values equal to 1 . Classification images were calculated individually for each participant and on the aggregate observer obtained by combining data from all participants . As a proof-of-principle to demonstrate that the classification images really represent the templates of prototypical streaming or bouncing events , we reverse engineered the stimuli , and used the classification images to create unambiguous stimuli . To do so , we first generated noisy stimuli like in Experiment 1 . Next , we added or subtracted the empirical luminance kernel of the aggregate observer of Experiment 1 ( Fig 2B ) to modulate luminance over time and space , and hence to generate disambiguated ‘bouncing’ and ‘streaming’ stimuli , respectively . The resulting stimuli ( S4 Movie ) provide a striking example of unambiguous dynamic displays: the “bouncing” stimulus ( S4 Movie , top ) is most likely perceived as bouncing , while the “streaming” stimulus ( S4 Movie , bottom ) is mostly seen as streaming . This was corroborated by showing the video to a pool of 12 naïve participants ( 3 female ) and asking them which of the two stimuli appeared to be streaming and which bouncing . As expected , all participants indicated the top stimulus as bouncing and the lower one as streaming . This further corroborates the validity of the present reverse correlation analyses , and phenomenologically demonstrates that the empirical classification images do indeed represent the templates for prototypical streaming or bouncing events . The extraction of visual sensory evidence Ev for perceptual classification is modeled in terms of total motion energy ( see below ) . In order to keep the model simple and because characterizing early stages of sensory processing is beyond the scope of the current study , we assumed the stimuli to be linearly transduced . The integration of task-relevant information E was modeled in terms of weighted linear summation of the form: Z ( bounce ) =ω0+∑iωiEi , ( 3 ) where ωi denotes the linear coefficient ( i . e . , the weight ) , Ei the evidence , ω0 the bias ( i . e . , the decision criterion ) , and the subscripts i the source of the evidence ( i . e . , visual motion energy , auditory click , previous response ) . An assumption of this model is that the internal noise for each task-relevant evidence Ei is independent and normally distributed . Such a linear model has one covariate Ev corresponding to the evidence from visual motion energy and two factors EA and ER ( t−1 ) denoting the presence or absence of a sound and the response given on the previous trial . The coefficients of the model were fitted individually for each participant and for the aggregate observer using the Matlab routine glmfit . Given that we were interested in understanding the effect of the previous trials on subsequent responses , the first trial of each block was not used for modeling purposes . The model was validated using a 39-fold cross-validation procedure ( for both individual and aggregate observers ) . The training set was used on each iteration to fit the coefficients of the model . Next , we fed the stimuli of the test set into the model and we compared the response of the model to the empirical data . Each trial was included in the test set in only one iteration . To evaluate how well the model could reproduce the observed responses , we partitioned all trials in 50 bins according to the model response Z ( bounce ) in the test set of the cross-validation . The predicted probability of reporting a bounce on each trial , p ( bounce ) , was calculated from the models’ response using the following equation: p ( bounce ) =Φ[Z ( bounce ) ] ( 4 ) where Φ[∙] is the cumulative normal distribution function . For each bin , we also calculated the observed probability of reporting a bounce and we plotted predicted vs . observed responses ( Fig 4B , S1 Fig right column ) . If the model accurately captures participants’ responses , data should lie along the identity line . The 99% confidence interval along the identity line was calculated based on the binomial distribution and the number of responses of each bin . Overall , the model could replicate observed responses with high accuracy . The current linear integration model was compared to alternative models generated by taking only a subset of the three predictors ( i . e . , motion energy , sound , or previous response ) or by also including interaction terms . The selected model outperformed all such alternatives , as assessed in terms of Akaike information criterion . The motion energy model [40] is a classic and biologically plausible model of visual motion detection based on the combination of a series of spatiotemporally tuned filters . Although a full description and rationale of the motion energy model can be found elsewhere ( e . g . , [40 , 56] , here we briefly describe its main features and provide the equations of the spatial and temporal filters . In the current study we implemented a recent version ( including the values of the relevant parameters ) of the motion energy model [56] . Motion filters had the same spatial and temporal extent as the visual stimuli and they were sampled at the same spatial and temporal resolution . The spatial filters consisted of even ( E ) and odd ( O ) Gabor functions: E ( x ) =cos ( 2πfx ) ∙e− ( xσ ) 2 ( 5 ) O ( x ) =sin ( 2πfx ) ∙e− ( xσ ) 2 . ( 6 ) The spatial constant σ was 0 . 5 deg and its spatial frequency f was 1 . 1 cpd . Such values approximate the spatial sensitivity of the magnocellular system [57] . Temporal filters were defined by the following equation: R ( t ) = ( kt ) n∙exp ( −kt ) ∙[1∕n ! −β ( kt ) 2∕ ( n+2 ) ! ] ( 7 ) The parameter k represents the center temporal frequency of the filter and its value was set to 100 . The parameter n represents the temporal constant of the filter and its value was set to 9 for the slow temporal filter and 6 for the fast one . The parameter β represents the weighting of the negative relative to the positive phase of the filter and its value was set to 0 . 9 . The model also includes a normalization step . A graphical representation of the full model is displayed in S4 Fig and a Matlab implementation of the model is available online ( http://www . georgemather . com/Code/AdelsonBergen . m ) . In the current study , the total motion energy of each stimulus was calculated . The motion energy matrix displayed in Fig 3 was obtained by computing the motion energy matrices from all the stimuli presented in the experiment , and averaging the results . Given that the results of Experiment 1 point to the role of contrast in perceptual disambiguation , we assessed whether a simpler model which is sensitive to contrast but not to motion is sufficient to explain the current results . For this , we implemented a model consisting of two biphasic spatial filters in quadrature pair with a Gaussian temporal profile . Such a model has been adopted , and fully described , in a related study which also relied on reverse correlation analyses [29] . The spatial filters ( and the values of the relevant parameters ) of this model are identical to those of the energy model presented here , and based on [29] we set standard deviation of the Gaussian temporal filter to 40ms . We used this alternative model to calculate the predicted classification image , just as we did for the motion energy model . S5 Fig shows that this simpler model , which does not respond to motion , simply cannot account for human performance . | Sensory information is generally ambiguous , and a single sensory modality most often cannot provide enough information to univocally specify the actual state of the world . A primary task for the brain is therefore to resolve perceptual ambiguity . Here we use a dynamic audiovisual ambiguous display embedded in noise to investigate the computational mechanisms of perceptual disambiguation . Results demonstrate that the brain first extracts visual information for perceptual disambiguation through motion detectors . Such information is next combined with auditory information–and memory of recent perceptual history–through weighted averaging to determine the final percept . This study revealed the particular spatiotemporal stimulus patterns that elicit a stream or a bounce percept , respectively , and it demonstrates that perceptual disambiguation is majorly affected by noise , whose random spatiotemporal patterns provide signal-like information that bias perception in a very systematic fashion . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"decision",
"making",
"engineering",
"and",
"technology",
"signal",
"processing",
"social",
"sciences",
"light",
"neuroscience",
"learning",
"and",
"memory",
"electromagnetic",
"radiation",
"perception",
"cognitive",
"psychology",
"luminance",
"cognition",
"memory",
"visible",
"light",
"vision",
"computer",
"and",
"information",
"sciences",
"physics",
"psychology",
"psychophysics",
"information",
"theory",
"biology",
"and",
"life",
"sciences",
"background",
"signal",
"noise",
"sensory",
"perception",
"physical",
"sciences",
"cognitive",
"science"
] | 2017 | Noise, multisensory integration, and previous response in perceptual disambiguation |
Viruses are usually thought to form parasitic associations with hosts , but all members of the family Polydnaviridae are obligate mutualists of insects called parasitoid wasps . Phylogenetic data founded on sequence comparisons of viral genes indicate that polydnaviruses in the genus Bracovirus ( BV ) are closely related to pathogenic nudiviruses and baculoviruses . However , pronounced differences in the biology of BVs and baculoviruses together with high divergence of many shared genes make it unclear whether BV homologs still retain baculovirus-like functions . Here we report that virions from Microplitis demolitor bracovirus ( MdBV ) contain multiple baculovirus-like and nudivirus-like conserved gene products . We further show that RNA interference effectively and specifically knocks down MdBV gene expression . Coupling RNAi knockdown methods with functional assays , we examined the activity of six genes in the MdBV conserved gene set that are known to have essential roles in transcription ( lef-4 , lef-9 ) , capsid assembly ( vp39 , vlf-1 ) , and envelope formation ( p74 , pif-1 ) during baculovirus replication . Our results indicated that MdBV produces a baculovirus-like RNA polymerase that transcribes virus structural genes . Our results also supported a conserved role for vp39 , vlf-1 , p74 , and pif-1 as structural components of MdBV virions . Additional experiments suggested that vlf-1 together with the nudivirus-like gene int-1 also have novel functions in regulating excision of MdBV proviral DNAs for packaging into virions . Overall , these data provide the first experimental insights into the function of BV genes in virion formation .
Microorganisms form associations with metazoan hosts that range from beneficial symbiosis ( mutualists ) to parasitic ( pathogens ) . Mutualists serve as important sources of evolutionary innovation for hosts , while pathogens often acquire genes from hosts or other organisms that facilitate their own survival and cause disease . Although most research on obligate mutualists focuses on bacteria , several fungi and protozoans are also known to form beneficial partnerships [1]–[3] . Viruses in contrast are almost always thought to form parasitic associations [4]–[6] . A notable exception to this is the family Polydnaviridae , which consists entirely of large DNA viruses that are obligate mutualists of insects called parasitoid wasps [7] , [8] . Polydnaviruses ( PDVs ) thus provide an opportunity for understanding the adaptations involved in the evolution of viruses into mutualists from pathogenic ancestors . Parasitoid wasps reproduce by laying eggs into other insects ( hosts ) that their progeny consume [9] . The Polydnaviridae consists of two genera: the Bracovirus ( BV ) associated with ca . 20 , 000 species of wasps in the family Braconidae , and the Ichnovirus ( IV ) associated with ca . 18 , 000 species of wasps in the family Ichneumonidae [10] . Each wasp species carries a genetically unique PDV that persists in all cells as an integrated provirus . Viral replication only occurs in pupal and adult stage female wasps in a type of cell in the ovary called calyx cells . Virions from calyx cells are released via cell lysis and accumulate to high density in the lumen of the reproductive tract to form calyx fluid . Virions are also enveloped and contain multiple , circular , double-stranded DNAs of large aggregate size ( 190–600 kbp ) that encode many virulence genes . Most PDV-carrying wasps parasitize larval stage Lepidoptera ( moths ) by depositing eggs containing the proviral genome plus a quantity of virions . These virions rapidly infect host cells , which is followed by expression of virulence genes that immunosuppress and alter the development of hosts in ways that are essential for survival of the wasp's progeny [11] . The origin and genomic organization of IVs is unclear [11] . In contrast , BV genomes exhibit features unlike any other known viruses [12]–[14] . As proviruses , their genomes consist of two types of DNA domains: those that contain genes with predicted roles in replication , and others that contain the virulence genes that become packaged into virions . Remarkably , these domains reside in different locations in the wasp genome [12] , [13] . In addition , while genes with predicted roles in replication are transcribed in calyx cells , their transmission is entirely vertical and independent of any viral DNA replication or encapsidation [7] , [11] , [12] . Virulence gene-containing domains are likewise transmitted vertically . However , they also are amplified , excised from the wasp genome into circular forms and packaged into virions during replication in calyx cells [7] , [15]–[20] . In all other cells of the wasp including the germ line both the replication and virulence genes of the proviral genome are silent [11] . BVs cause no apparent disease in wasps because almost no virulence genes are expressed in wasp cells and lytic replication is restricted to only calyx cells [11] , [21] . In contrast , BVs cause severe disease in the hosts wasps parasitize because virions systemically infect the host insect and all of the virulence genes virions deliver are expressed [11] , [21] , [22] . The disease symptoms BVs cause in the host , however , are also essential for development of the wasp's offspring . Thus , BVs depend on wasps for genetic transmission , while wasps depend on BVs for parasitism of hosts . Genes in the proviral genome of BVs with predicted roles in replication were identified because they exhibit homology with core genes from two other types of arthropod-infecting viruses: nudiviruses and their sister taxon the Baculoviridae [7] , [12] . Like BVs , nudiviruses and baculoviruses replicate in cell nuclei and package large , circular ds-DNA genomes into enveloped virions . Unlike the developmentally-linked and tissue-specific replication of BVs , however , baculoviruses are virulent pathogens , which establish systemic infections in insects by undergoing lytic replication in virtually all cells of the infected host and expressing a variety of virulence genes [23] . Nudiviruses produce either systemic , lytic infections or latent infections [23] , [24] . More than 60 baculoviruses have been sequenced and a survey of a subset ( 13 ) of these genomes indicates that all share 31 genes , which are collectively referred to as the baculovirus core gene set [24] , [25] ( Figure 1 ) . Functional studies of model species like Autographa californica multinucleopolyhedrosis virus ( AcMNPV ) indicate about half of these genes have essential roles in viral replication [24] . These include genes with roles in replicating the viral genome , several genes that code for virion structural components , and four genes that code for subunits of a novel RNA polymerase , which selectively transcribes viral genes because it recognizes unique promoter sequences ( Figure 1 ) [24] . Six nudivirus genomes have been sequenced and each contains 20 genes with homology to structural , replication and transcription components of the baculovirus core gene set [23] ( Figure 1 ) . The actual function of these genes , however , is unknown beyond inferences from baculoviruses . Data from three braconid wasps , Cotesia congregata , Chelonus inanitus , and Microplitis demolitor , indicate they lack recognizable homologs of most baculovirus core genes with roles in viral DNA replication ( Figure 1 ) , which suggests that , unlike baculoviruses , replication of BV DNAs packaged into virions is regulated by machinery from the wasp [7] . However , BVs do encode homologs of several baculovirus/nudivirus-like structural genes plus the four subunits of a baculovirus/nudivirus-like RNA polymerase [7] , [12] . Each of these genes is also transcribed in ovaries when BV virions are produced . Together , these genes form a conserved gene set likely present in all BV genomes ( Figure 1 ) . However , we refrain from referring to these as “core” genes because of the small number of BV genomes currently available and their non-discrete organization in wasp genomes [7] . Other predicted members of a conserved BV gene set include a baculovirus/nudivirus-like sulfhydryl oxidase ( ac92 ) , 11 nudivirus-like genes unknown from baculoviruses , and 11 novel genes [7] . Since BV-carrying braconids are monophyletic [26] , these data overall indicate that BVs evolved from an ancestral nudivirus-wasp association . Fossil calibrations estimate this association arose 100 million years ago ( Mya ) , while the last common ancestor of BVs , nudiviruses , and baculoviruses existed approximately 312 Mya [27] . Given these timelines and the pronounced differences that exist today between BVs and baculoviruses , it is not surprising many of the genes they share have diverged to the point that homology is difficult to recognize outside of essential residues or functional domains . Indeed , algorithms like BLAST cannot detect homology between BV and baculovirus genes , while identity between predicted BV and more closely related nudivirus proteins ranges between 19–41% [7] . Such divergence , however , also begs the question of whether baculovirus-like genes in BV proviral genomes retain baculovirus-like functions . Here , we used proteomic , RNA interference ( RNAi ) , and functional assays to characterize selected members of the conserved gene set of Microplitis demolitor bracovirus ( MdBV ) in the wasp M . demolitor . Our results indicate that six genes with hypothesized roles in replication exhibited conserved functions relative to baculoviruses . Our data also identified novel functions for two genes in excision of viral DNAs for packaging into virions .
BV replication in calyx cells begins with amplification of a portion of the proviral genome , which is followed by the de novo assembly and packaging of virions in nuclei [20] , [28] . Calyx cells then lyse which releases virions into the lumen . In the case of MdBV , prior studies establish the timing of these events and the chronology of replication gene expression during the pupal and adult stages of M . demolitor [7] . MdBV packages multiple circular , double-stranded DNA segments into virions but each individual virion contains only a single viral DNA [14] , [29] . The sequence of these DNAs as episomes and their wasp-viral boundary sequences when integrated into the genome of M . demolitor are known [14] , [29] , [30] . Prior studies document that these DNAs are specifically amplified in M . demolitor calyx cells , followed by excision from flanking DNA and circularization [7] . Flanking wasp DNA at the site of excision is then rejoined to form an ‘empty locus’ [7] . Given this background , we first conducted a proteomic analysis of MdBV virions to determine whether predicted conserved structural components were present . To accomplish this , we produced two independent samples of calyx fluid with the second sample further purified on a sucrose gradient that produces morphologically pure and intact virions [31] . Following separation on SDS-PAGE gels , proteins were in-gel trypsin digested and analyzed using an Orbitrap Elite mass spectrometer . Mass spectra were then compared to our previously generated M . demolitor ovary transcriptome database [7] to identify proteins present . We present our findings relative to the predicted core/conserved gene sets for baculoviruses , nudiviruses , and BVs in Figure 1 and Table S1 . Four proteins ( 38K , VP39 , VLF-1 , AC92 ) detected in MdBV virions were homologs of baculovirus capsid or capsid/envelope components . VP39 was the most accurately detected of these proteins based on the total number of unique peptides identified , which corresponded with vp39 also being the most abundant viral gene transcript detected in ovaries during MdBV replication [7] ( Table S1 ) . We also detected several proteins related to envelope components of baculovirus occlusion-derived virus . These included variants of ODV-E66 and ODV-E56 ( = PIF-5 ) plus the infectivity factors P74 , PIF-1 through -6 and envelope component VP91 ( Figure 1 , Table S1 ) . Seven MdBV virion proteins corresponded to genes in the BV conserved gene set for which homologs are known from all or some nudiviruses but are unknown from baculoviruses ( Figure 1 ) . These included the product of the integrase-1 ( int-1 ) gene , which is structurally related to vlf-1 , plus products of several nudivirus-like genes of unknown function ( HzNVORF9-1 and -2 , 64 , 94 , 106 , PMV Hypothetical Protein ) . We also detected products of four conserved genes or gene families unique to BVs ( 17A , 35A , 97A ) ( Figure 1 , Table S1 ) . In contrast , we did not detect any proteins in virions that corresponded to the helicase gene or the RNA polymerase subunits ( lef-4 , lef-8 , lef-9 , p47 ) ( Figure 1 ) . The preceding data showed that MdBV virions contain BV conserved gene products but provided no experimental evidence for their function . We therefore selected six genes from MdBV for functional studies . Our choices included two predicted subunits of a baculovirus-like RNA polymerase ( lef-4 , lef-9 ) , 2 ) , two homologs of baculovirus capsid genes ( vp39 , and vlf-1 ) , and 3 ) two homologs of baculovirus envelope genes ( p74 and pif-1 ) . Each of these genes is a member of the BV conserved gene set because orthologs are likely present in all other BVs studied to date ( see Figure 1 ) . Each is also a member of the baculovirus core gene set with prior studies from AcMNPV or other isolates providing experimental evidence for the function of each [24] . In addition , we selected one nudivirus-like gene ( int-1 ) , unknown from baculoviruses , for which no functional studies have been conducted . As previously noted , sequence divergence between members of the BV conserved gene set and corresponding baculovirus/nudivirus core genes is high . Identities of the above gene products from MdBV with corresponding predicted proteins from the closest known nudivirus relative , Heliothis zea nudivirus-1 ( HzNV-1 ) , were: lef-4 ( 25% ) , lef-9 ( 32% ) , vp39 ( 19% ) , vlf-1 ( 28% ) , p74 ( 26% ) , pif-1 ( 28% ) , and int-1 ( 30% ) . The genes we selected reside in the MdBV proviral genome and each is transcribed in ovary calyx cells during replication [7] . However , conventional knock out techniques used to characterize baculovirus gene function are untenable because the DNA domains where these genes reside are not replicated and packaged into MdBV virions . We thus assessed whether RNAi could be used to knock down transcription of these genes in M . demolitor . Since MdBV replication begins in the pupal stage of the wasp , we developed methods for injecting gene-specific dsRNAs into wasp larvae after they emerged from a host caterpillar and spun a cocoon . We then compared the abundance of each targeted transcript in newly emerged adult wasps by qPCR relative to treatment with a non-specific dsRNA ( ds-eGFP ) . Our results showed that we reduced transcript abundance on average 60–99% for each gene we targeted ( Figure 2 ) . Using an antibody we generated to MdBV LEF-9 , we also confirmed that knockdown at the transcript level resulted in knockdown of the corresponding protein ( Figure 2 ) . Before initiating any functional experiments , we further verified our approach by examining the effects of dsRNA dose , time required after treatment for target transcript degradation , and specificity . Using vlf-1 as an example , our results showed that injection of 50 ng to 5 µg of dsRNA per wasp larva yielded a similar level of knockdown ( Figure S1A ) . Our results also indicated that injection of vlf-1 dsRNA into wasp larvae did not significantly reduce transcript abundance until day 3 of the pupal stage , which suggested that 2–3 days were required before an RNAi effect was observed ( Figure S1B ) . We examined specificity of knockdown in two ways . Since vlf-1 and int-1 are homologous genes , we verified that int-1 dsRNA , which strongly knocked down int-1 ( Fig . 2G ) , did not affect transcript abundance of vlf-1 via off-target effects [32] ( Figure S2A ) . We also generated a second vlf-1 dsRNA that did not overlap the dsRNA used for the data presented in Figure 2D to verify that it had a similar knockdown effect on vlf-1 transcript abundance ( Figure S2B ) . Baculovirus RNA polymerases consist of four subunits ( LEF-4 , LEF-8 , LEF-9 , P47 ) , which transcribe baculovirus genes with roles in virion formation [33] . These subunits are categorized as ‘early’ genes because they are transcribed before DNA replication and transcription of structural ‘late’ and ‘very late’ genes commences . As noted above , baculovirus RNA polymerases selectively transcribe late and very late viral genes because they recognize unique promoter elements with the consensus sequence ( A/G/T ) TAAG absent from host insect genes [34] , [35] . The conserved gene set of BVs contains homologs of each RNA polymerase subunit . Expression data from M . demolitor and Cotesia congregata also indicate these subunits are transcribed in ovaries earlier than predicted structural genes , while sequence analysis has identified baculovirus late gene promoter elements upstream of the start codon of several predicted BV structural genes [7] , [12] . Thus , if the BV RNA polymerase subunits form a functionally similar enzyme as baculoviruses , RNAi knockdown of one or more subunits should compromise transcription of MdBV structural genes but not wasp genes . As shown above ( Figure 2A , B ) , we knocked down lef-4 , a predicted 5′ capping enzyme [36] , [37] and lef-9 , a predicted RNA polymerase subunit that forms part of the catalytic cleft [38] , [39] . We then measured transcript abundance of two predicted MdBV structural genes ( vp39 and p74 ) , and two typical insect genes expressed in ovaries ( elongation factor 1 alpha ( ef1α ) and DNA polymerase delta subunit ( dnapolδ ) ) . Our results showed that knockdown of lef-4 and lef-9 significantly reduced transcript abundance of vp39 and p74 while having no effect on ef1α and dnapolδ ( Figure 3A–D ) . Given this outcome and the putative role of vp39 and p74 as structural genes we reasoned that reduced expression of vp39 and p74 could also result in production of fewer virions on average than control females . We therefore estimated viral titer by using episomal MdBV DNA segment B as a marker and a previously developed qPCR assay that includes a DNAse step to remove all non-encapsidated DNA before isolating DNA from ovary homogenates [7] . In this manner , the copy number of episomal segment B in virions , which protect the packaged DNA , could be determined . These results showed that knockdown of lef-4 and lef-9 significantly reduced the titer of DNAse-protected segment B relative to ds-eGFP-treated controls ( Figure 3E ) . In baculoviruses , VP39 is a major capsid protein while VLF-1 is a structural component , and is also functionally required for capsid production and very late gene expression [24] , [40]–[42] . After knocking down vp39 and vlf-1 in M . demolitor ( Figure 2C , D ) , we first assessed whether either treatment affected virion structural integrity by measuring the DNase sensitivity of packaged genomic DNAs as described above . These assays indicated that the abundance of DNase-protected segment B declined by 83% and 78% in vp39 and vlf-1 knockdown samples respectively relative to the control ( Figure 4A , B ) . We also used the non-overlapping dsRNA , ds-vlf-1-2 in these assays , which produced the same result as ds-vlf-1 ( Figure S2C ) . We then examined the effects of vp39 and vlf-1 knockdown on the ability of MdBV to infect cells from the moth Chrysodeixis includens , which is a host for M . demolitor . For these assays , we used CiE1 cells , which is a continuous , hemocyte-like cell line established from C . includens that is highly permissive to MdBV infection [30] , [43] . We determined by qPCR that 2–4 copies of episomal DNA segment B were present per CiE1 cell when cultures were infected at an estimated MOI of 100 with MdBV from control wasps treated with ds-eGFP ( Figure 4C , D ) . In contrast , copy number was 85 . 2% and 69 . 6% less when cells were infected with the same amount of calyx fluid from vp39 and vlf-1 knockdown wasps . We also assessed infection using the MdBV gene product GLC1 . 8 , which is an excellent marker because it is rapidly expressed on the surface of CiE1 cells and is easily visualized immunocytochemically [43] , [44] . Normalizing the control samples , we observed that vp39 knockdown reduced the fraction of cells stained for GLC1 . 8 to less than 10% , while vlf-1 knockdown reduced this fraction to 26 . 8% ( Figure 4E , F ) . These results could be explained by vp39 and vlf-1 knockdown either adversely affecting virion formation , which would result in calyx fluid containing a lower titer of virus , or causing structural defects that do not reduce virion density but nonetheless compromise function . We therefore examined virion morphology in calyx fluid by transmission electron microscopy ( TEM ) . We previously documented that MdBV virions in calyx fluid consist of a single barrel-shaped nucleocapsid surrounded by a highly elongate envelope [29] , [31] . By counting the number of virions in randomly selected fields of view from treatment and control wasp sections , we determined that calyx fluid from a vlf-1 knockdown wasp contained a slightly lower concentration of virions than observed in a control wasp , whereas a vp39 knockdown wasp did not ( Figure 5A–D ) . We then examined MdBV morphogenesis in calyx cell nuclei . Early studies of BVs show that calyx cells exhibit a progression of development with smaller , younger cells being situated closer to the ovarioles and older , large cells being closer to the lumen of the ovary [45] . In turn , young cells show no evidence of BV replication , while old calyx cells contain an abundance of assembled virions in their nuclei [45] , [46] . In control wasps , we observed that MdBV morphogenesis began with the de novo appearance of short membrane profiles in calyx cell nuclei . This was followed by the formation of nucleocapsids near virogenic stroma , which is where DNA packaging also occurs in baculoviruses . Assembled MdBV virions then formed large aggregations with a layered crystalline structure ( Figure 5E ) . At this stage , virions were rod-shaped and of uniform length . The envelope surrounding each nucleocapsid was also not as elongated as seen for virions in calyx fluid ( see Figure 5A versus 5E ) . Calyx cells from vlf-1 knockdown wasps in contrast exhibited an abundance of membrane profiles that appeared to be envelope progenitors ( Figure 5F ) . Rather than elongating , these envelopes were spherical and either lacked capsids entirely or contained empty capsids ( Figure 5F ) . A number of electron dense and empty capsids were also observed with no envelope ( Figure 5F ) . Lastly , almost no aggregations of rod-shaped , electron dense virions were present in calyx cells from vlf-1 knockdown wasps . Calyx cells from vp39 knockdown wasps showed no distinct alterations in virion assembly , but aggregations of rod shaped , electron dense virions were consistently much smaller than those observed in control wasps ( Figure 5G ) . p74 and the pif genes are known as per os infectivity factors because their loss in baculoviruses such as AcMNPV disables oral infection of host insects by occlusion derived virus [47]–[49] . Each is also a component of the occlusion derived virus envelope where they form a complex with one another [50] . Unlike baculoviruses , BV virions never infect host insects orally but instead are injected into the hemocoel by wasps where they bind to host cells such as hemocytes via fusion of the envelope with the plasma membrane [51] . Nucleocapsids then travel through the cytoplasm to nuclear pores where they release their DNA into the nucleus to initiate transcription of virulence genes like glc1 . 8 [51] , [52] . Given the differences in the known functions of per os infectivity factors in baculoviruses relative to the biology of BVs we asked whether the p74 and pif-1-like genes of MdBV still play a role in infectivity by knocking down each ( Figure 2E , F ) and then conducting the same assays in CiE1 cells as described above . Our results revealed no differences between knockdown and control wasps in the copy number of DNA segment B in CiE1 cells at 24 h post-infection ( Figure 6A ) . However , the fraction of CiE1 cells expressing GLC1 . 8 on their surface was dramatically lower using virus from p74 and pif-1 knockdown wasps ( Figure 6B ) . As noted above , all baculoviruses encode a vlf-1 gene while nudiviruses and BVs also encode related integrase genes ( known as vlf-1 or vlf-1a , vlf-1b-1 and -2 or HzNVORF140 , and int-1 and -2 or HzNVORF144 ) [7] , [12] . Although VLF-1 is a structural component of baculovirus virions , it along with nudivirus integrase genes are members of the tyrosine ( Tyr ) recombinase family , which includes several enzymes that mediate the excision and integration of genetic elements [53] . As noted above , elimination of vlf-1 from AcMNPV disables capsid formation while mutation of the conserved Tyr residue required for integrase activity in other Tyr recombinase family members produces non-infectious virus [40] . Whether baculovirus VLF-1 possesses any integrase activity , however , remains unstudied in all likelihood because baculovirus genomes persist as episomes in infected host cells and are unknown to integrate . In contrast , a key feature of BVs is their persistence in wasps as integrated proviruses that amplify , excise and package a portion of the genome when replicating in calyx cells . We therefore assessed whether vlf-1 and/or int-1 homologs from MdBV regulate proviral DNA excision . We had previously determined that the MdBV proviral genome encodes three distinct vlf-1 genes ( named vlf-1 , vlf1b-1 , -2 ) and two integrase genes ( int-1 , -2 ) that are all transcribed in ovaries during replication [7] . Phylogenetic analysis further suggested the integrase genes of BVs likely arose from duplication of vlf-1 in the nudivirus ancestor , which was then followed by duplication of each gene in M . demolitor [7] . Alignment of vlf-1 and integrase family members from MdBV , select nudiviruses , AcMNPV , and Chelonus inanitus bracovirus ( CiBV ) showed that MdBV vlf-1 and int-1 both retain a typical active site Tyr residue for predicted integrase activity , whereas other M . demolitor family members do not ( Figure 7A ) . We thus knocked down vlf-1 and int-1 ( Figure 2D , G ) , and then isolated DNA from newly emerged adult wasp ovaries to determine whether proviral DNAs had excised from the wasp genome as normally occurs . This was accomplished using MdBV segment B as a marker and qPCR assays that measured copy number of the rejoined ‘empty locus’ that only forms if proviral DNA segment B was excised from the wasp genome ( Figure 7B ) . Copy number of the empty locus in ovaries from control wasps treated with ds-eGFP was 14 . 4×106 , which indicated a high level of excision of DNA segment B from calyx cells . In contrast , we detected almost no copies of the empty locus in wasps treated with ds-int-1 and ds-vlf-1 ( Figure 7C ) .
Phylogenetic data strongly support that BVs evolved from a nudivirus ca . 100 Mya , and that nudiviruses and baculoviruses shared a more ancient common viral ancestor ca . 200 Mya earlier [12] , [26] , [27] . Detailed studies of AcMNPV and select other species also provide important insights into the function of baculovirus core genes . In contrast , the hypothesized function of baculovirus core gene homologs in BVs ( and nudiviruses ) is founded on inferences from the baculovirus literature and/or expression patterns in wasp ovaries during replication . Thus , the primary goal of this investigation was to assess whether RNAi methods could be used to disrupt BV gene expression , and then to use these methods with a subset of genes to determine whether their roles in replication were consistent with or differed from baculoviruses . Prefacing these functional studies , we conducted a proteomic analysis of purified MdBV virions to assess whether BV conserved genes that are homologs of baculovirus structural components were present . We also did this to compare MdBV virions to virions from CcBV and CiBV , which are the only other BVs for which any proteomic data are available [12] , [54] . We detected all of the baculovirus-like capsid and envelope components previously identified in CcBV and CiBV as well as two additional baculovirus-like conserved genes not detected in CcBV or CiBV . These included AC92 , which is associated with baculovirus nucleocapsids , and PIF-3 , which is associated with baculovirus occlusion-derived virus envelopes [55] . We also detected four nudivirus-like ( HzNVORF9-1 and -2 , HzNVORF106 , PmV ) and three novel conserved gene products ( 17A , 35A , 97A ) in MdBV virions identified in CiBV virions . In contrast , we did not detect three novel gene products ( 27B , 30B , 97B ) reported from CiBV , yet did detect three nudivirus-like gene products ( INT-1 , HZNVORF64 , HZNVORF94 ) not reported from CiBV virions . Proteomic data must be interpreted cautiously when investigating viral structural proteins , because of the potential for non-integral proteins to become non-specifically associated with virions during assembly [24] . Nonetheless , the baculovirus literature combined with our detection of the capsid and envelope proteins in Figure 1 suggest these baculovirus-like gene products are likely structural components of MdBV virions . While no functional data from nudiviruses exist , we speculate for the same reason that products of the nudivirus-like conserved genes HzNVorf 9-1 , 9-2 , -64 , -94 , -106 , PmV hypothetical protein and novel conserved genes 17a , 35a , and 97a , are also structural proteins . Our proteomic data did not identify any peptides corresponding to MdBV conserved genes with predicted roles in DNA replication ( helicase ) or transcription ( lef-4 , lef-8 , lef-9 , p47 ) , which at minimum indicates our samples were not contaminated with some non-integral products transcribed in calyx cells [7] . However , it is notable that we consistently detect the products of the int-1 , vlf-1b-1 and vlf-1b-2 genes , which could suggest that similar to baculovirus vlf-1 they too are capsid components or are packaged into capsids together with episomal DNA . Because of the unique biology of BVs and limited genetic data available for their associated wasps , the options available for studying gene function are obviously constrained . RNAi is a potentially powerful method for studying BV gene function , but its efficacy in insects is also patchy with examples of successful use being more prevalent in some taxa ( beetles ( Coleoptera ) , mosquitoes ( Diptera ) ) [56] , [57] than others ( moths ( Lepidoptera ) [58] ) . We thus were very careful in validating our RNAi approach for knocking down MdBV genes in M . demolitor before initiating any functional studies . Our analysis of the ovary transcriptome indicated that all genes of the siRNAi pathway are present in M . demolitor and transcribed [7] . Results presented in this study further show that ds-RNA injection into late larval stage M . demolitor effectively and specifically knocks down the genes we targeted . While we present the outcome of a number of validation experiments using vlf-1 as an example , we have conducted similar experiments with other MdBV conserved genes , which all showed the same trends . With knockdown methods established , we used the baculovirus literature and our proteomic data to select six genes in the MdBV conserved gene set with hypothesized roles in viral transcription ( lef-4 , lef-9 ) , capsid assembly ( vp39 , vlf-1 ) , and envelope formation ( p74 , pif-1 ) . Our rationale for selecting these genes was also driven by the strength of the functional literature for each in baculoviruses , which provided in most cases clear expectations for what a conserved phenotype should be for MdBV . Our results with lef-4 and lef-9 strongly support that MdBV produces a baculovirus-like RNA polymerase that preferentially transcribes structural genes . We also note that knockdown of lef-4 more strongly disabled structural gene expression than knockdown of lef-9 . This could reflect that as a capping enzyme the effect of knocking down of lef-4 was further enhanced by degradation of cap-lacking transcripts . The transcription of reporter virus structural genes vp39 and p74 was not completely abolished for either lef-4 or lef-9 knockdowns . Despite detecting no LEF-9 protein after knockdown on immunoblots , this could reflect incomplete knockdown , and the presence of enough RNA polymerase subunit proteins to produce some functional viral RNA polymerase holoenzyme . Alternatively , viral structural genes may also be transcribed in part by wasp RNA polymerase II . Activity of replication gene transcription compared to relative silence of BV genes that are ultimately packaged into virions suggests that replication genes are transcribed from ancestral viral RNA polymerase promoters whereas virulence genes are not . The phenotypic effects we observe in response to vlf-1 knockdown are consistent with this protein being both a structural component and a product required for virion assembly . The defects in morphogenesis of MdBV virions we observe , however , differ somewhat from the defects observed with AcMNPV where knockout of vlf-1 resulted in formation of tubular structures that appeared to be aberrant capsids that fail to package DNA [40] . The technical approaches to these studies , however , resulted in observations associated with budded virus production and precluded examination of potential defects associated with formation of occlusion-derived virus ( see below ) . In contrast , the severe defects we observed in the assembly of MdBV virions suggest vlf-1 may be important in both DNA packaging and proper association of capsids with envelopes . Like vlf-1 , knockdown of vp39 greatly increased the sensitivity of packaged DNA to DNAse treatment while also reducing infectivity . For both genes , dsRNA treatment did not completely abolish DNAse protection or infectivity of virus particles , which we presume is due to incomplete knockdown of transcript levels . Unlike vlf-1 though , knockdown of vp39 did not cause any obvious morphological defects in virion assembly with the exception that virion aggregations in calyx cells were much smaller owing possibly to a reduction in VP39 for production of capsids . Given these observations , we are unclear why knockdown of vp39 did not reduce virion density in calyx fluid . Unlike BVs , which produce only one virion type , AcMNPV and most other baculoviruses produce two virion phenotypes named occlusion-derived virus and budded virus [24] . Occlusion-derived virus is embedded in a protein matrix called an occlusion body and is the type that initiates a midgut infection when ingested by a new host . Budded virus in contrast is non-occluded and is the form of the virus that disseminates from the midgut and other cells to systemically infect the insect . Occlusion-derived and budded virus capsids contain the same structural proteins [55] but their envelopes differ greatly with the former assembling de novo in host cell nuclei and containing products of several core genes including per os infectivity factors . Budded virus in contrast acquires an envelope when budding through host cell plasma membranes , which contains only one or two viral proteins ( GP64 , F ) [55] . Although never occluded , the de novo assembly of BVs in calyx cell nuclei together with the envelope components detected in their virions ( see above ) indicate that BV particles structurally share more features with the occlusion-derived phenotype of baculoviruses . On the other hand , while per os infectivity factors are required for infectivity and binding of the occlusion-derived phenotype of AcMNPV to midgut cells , they are not required for infection of cultured cells or host larvae when injected into the hemocoel [49] , [59] . We thus were unclear what effect , if any , knockdown of p74 or pif-1 might have on infectivity given that BVs infect host insects only when injected into the hemocoel by wasps . Our results reveal no defects in the copy number of MdBV DNA detected in CiE1 cells after infection with a high MOI . Yet , knockdown of each gene resulted in a large decline in the fraction of infected cells that expressed the marker gene GLC1 . 8 . These findings are interesting because they suggest the loss of per os infectivity factors from the MdBV envelope results in improper translocation of MdBV to host cell nuclei where transcription of glc1 . 8 and other virulence genes occurs . No such activity has previously been associated with per os infectivity factors in baculoviruses but intriguingly GP64 , the envelope fusion protein of budded virus , has been implicated in affecting baculovirus translocation to host cell nuclei [60] . In addition to targeting six baculovirus-like conserved genes , we also examined the function of nudivirus-like int-1 because this gene and vlf-1 are both tyrosine recombinase family members and BV replication requires the excision of proviral genomic DNAs from the wasp genome for packaging into virions . The near complete inhibition of empty locus formation after vlf-1 and int-1 knockdown suggests a role for both in proviral DNA excision . At this time we have little understanding of the recombination reactions VLF-1 and INT-1 potentially mediate , although detailed investigation of other tyrosine recombinases suggest they perform recombination events by establishing a synapse between their cognate binding sites and performing two consecutive strand exchanges [61] , [62] . If on the same molecule recombination between two binding sites leads to excision of a circular intermediate [63] . Analysis of tyrosine-recombinase structural properties also suggest the mechanism of recombination requires binding of four enzyme monomers , which suggests the possibility for involvement of both vlf-1 and int-1 in proviral DNA excision [53] . Additionally , proviral DNA segments that become encapsidated possess direct repeats at the sequence boundaries that abut the flanking wasp DNA , which have been previously hypothesized to contain binding sites for recombinases that mediate excision [18] , [19] , [30] . Lastly detection of both VLF-1 and INT-1 in virions suggests these factors may also have important functions in parasitized hosts given recent findings that all DNA segments packaged into MdBV virions rapidly integrate into the genome of infected host cells [30] . Taken together , our results provide the first experimental insights into the function of a subset of MdBV genes . We fully recognize that additional experiments will be needed to more fully characterize the function of the individual genes we targeted , but by examining several key genes at once we provide evidence that: 1 ) RNAi can be used to knock down a number of MdBV genes , and 2 ) the baculovirus-like genes we targeted exhibit conserved functions despite divergence from nudiviruses more than 100 Mya . At the same time our results with vlf-1 and int-1 also identify novel functions unknown from baculoviruses but essential to the biology of BVs . With these results in hand , we are now positioned to undertake both more detailed experiments on these genes as well as studies on nudivirus-like and novel genes in the BV conserved gene set for which expectations about function are less clear . The arms race between wasps and the hosts wasps parasitize likely underlies the high rates of speciation of BV-carrying braconids [26] , [64] . The genetic mechanisms guiding host range evolution in contrast are largely unknown . One hypothesis would be that PDV virion structure has undergone rapid adaptation in response to the different lepidopteran host species each wasp species parasitizes , which could result in high variation in BV virion structure . The similarities thus far found in BV conserved genes together with the functional insights provided here , however , strongly suggest that BV gene functions will be conserved across isolates . Thus , differences in the sequence of BV genes may affect whether a given isolate can infect a given host species but we think it unlikely large differences will be found in the structure of BV virions across isolates . In contrast , the literature already indicates that the virulence genes BVs package into virions vary greatly among isolates associated with phylogenetically distant species of wasps . Thus , we would expect that differences in the virulence genes BVs deliver to hosts strongly affect host range by impacting the ability of wasp offspring to successfully develop . Finally , conservation in the function of the MdBV RNA polymerase and structural proteins suggest that key features in the evolution of BVs into mutualists do not involve radical changes in virion structural products but rather in how: 1 ) transcription of early factors required for DNA replication and viral transcription are regulated so that replication only occurs in calyx cells , 2 ) the genome is organized so that only some portions are amplified and packaged into virions , and 3 ) virulence genes are kept silent in wasps .
All studies were approved by the Biological Safety and Animal Care and Use Committee of the University of Georgia and were performed in compliance with relevant institutional policies , National Institutes of Health regulations , Association for the Accreditation of Laboratory Animal care guidelines , and local , state , and federal laws . M . demolitor parasitizes several species of larval stage Lepidoptera including Chrysodeixis ( = Pseudoplusia ) includens . Both species were reared at 27°C with a 16 h-light:8 h-dark photoperiod . M . demolitor has an 11 day developmental period described in detail elsewhere [7] . For this study , M . demolitor females were allowed to parasitize C . includens larvae , and wasp offspring were then allowed to develop in the host for 6 days . On day 7 , wasp larvae emerge from hosts and spin a cocoon within several hours . Cocoons are slightly asymmetrical with the anterior end generally being more elevated and pointed than the posterior end that contains the wasp abdomen . Nine-12 hours after spinning their cocoon , wasps pupate and develop for four more days until emerging as an adult . Adult wasps were then maintained in constant dark at 18°C . Virus was collected from M . demolitor ovaries . For the first replicate , 100 whole ovaries were crushed in PBS and the debris was removed by centrifugation . Then , the supernatant containing the virus was spun down at 20 , 000×g for 5 minutes and washed with PBS 3 times to collect virus particles . For replicate 2 , calyx fluid was dissected from 100 dissected ovaries and resuspended in PBS . Virions were then isolated on a sucrose gradient as previously described [31] . Both virus samples were electrophoresed on either a 4–20% or 12 . 5% Tris-Glycine gel ( Lonza ) . For each sample , the entire lane was cut into four pieces to separate proteins by size . In-gel trypsin digestion was performed for each gel slice , by overnight incubation with trypsin ( 20 µg/ml ) in 20 mM ammonium bicarbonate . Tryptic fragments were extracted with 50% acetonitrile and 0 . 1% TFA and vacuum dried . Samples ( 4 µl ) were analyzed by an Orbitrap Elite mass spectrometer , coupled to an Easy-nLC II Liquid Chromatography ( LC ) instrument ( Thermo Fisher Scientific ) . Samples were desalted and pre-concentrated on a C18 Easy LC pre-column ( 100 um internal diameter ( ID ) ×2 cm , 5 µm particle packing ( PP ) ) . Peptides were eluted from a reverse-phase column ( 75 um ID ×10 cm , 3 µm PP ) with a gradient of 10–35% B for 70 min , 35–95% B for 10 min , 95% B for 5 min ( A = 0 . 1% formic acid in water , B = 0 . 1% formic acid in acetonitrile ) at 300 nl/min . Nanospray ionization was performed with a spray voltage of 2 kV , with a capillary temperature of 200°C . The Orbitrap mass analyzer was used to provide resolutions of 120 , 000 and 30 , 000 for MS and MS/MS analyses , respectively . Briefly , a cycle of one full-scan mass spectrum ( 300–2000 m/z ) was performed , followed by continuous cycles of CID and HCD MS/MS spectra acquisitions of the 2 or 5 most abundant peptide ions throughout the LC separation until the candidate ions were exhausted . Data were acquired using Xcalibur software ( v2 . 2 , Thermo Fisher Scientific ) . Proteins were identified by searching against a custom database “Md” consisting of translated open reading frames ( ORFs ) greater than 33 amino acids in size from transcripts described in [7] using the Mascot v2 . 3 algorithm ( Matrix Science Inc . ) . Transcripts can be accessed through NCBI accession numbers JO913492 through JO979916 and JR139425 through JR139430 . Data were visualized with ProteomeDiscoverer v1 . 3 ( Thermo Fisher Scientific ) . Peptides with scores greater than the identity score ( p<0 . 05 ) were considered significant matches . Only ORFs that were matched by at least two peptide spectra were considered positive identifications . To target individual genes for RNAi knockdown , gene-specific primers were designed with added T7 promoter adaptors to amplify a 300–400 bp template for double-stranded RNA ( dsRNA ) synthesis ( Table S2 ) . cDNA from adult wasp ovaries was amplified using standard PCR and the resulting products were used as template in the MegaScript RNAi Kit ( Ambion ) . Larval stage wasp cocoons were marked within 15 h of spinning and were subsequently injected within 1–3 h . As M . demolitor is protandrous with males developing faster than females , the cocoons selected for injections were biased towards later emergence times and thus female wasps . Cocoons were affixed to double-sided tape , and oriented so that the posterior ends were facing in the same direction . Cocoons were pierced with a minuten pin in the abdomen region and wasps were injected through the cocoon with a glass needle directly into the abdomen . Approximately 0 . 5–1 µl of 2–4 µg/µl dsRNA was injected into each individual . All control wasps were injected with a non-specific dsRNA probe homologous to the bacterial eGFP gene . Wasps were sampled within 24 h of emerging as adults , and ovaries were removed and separated using opthalmic scissors . One ovary half was snap-frozen at −80°C for RNA extraction , and the other was used to assay RNAi phenotypes as described below . Extraction of total RNA was performed using the QIAGEN RNeasy kit following the standard kit protocol with a 20 min on-column DNAse treatment and elution in 30 µl of RNAse-free H2O . RNA concentration was measured using a Nanodrop Spectrophotometer and cDNA was synthesized from a quantity of total RNA normalized across samples . First-strand cDNA synthesis was performed using Invitrogen reagents including the Superscript III enzyme and oligo ( dT ) primers as outlined by the manufacturer ( Invitrogen ) . We used quantitative PCR ( qPCR ) to detect differences in expression of target genes after dsRNA treatment . An absolute standard curve was generated via PCR amplification of the corresponding cDNA for each gene of interest using specific primers ( Table S2 ) . Each product was cloned into pSC-A-amp/kan , and after propagating and isolating each plasmid from minipreps , its identity was confirmed by sequencing . Standard curves were generated followed by determination of copy numbers from serially diluted amounts ( 102 to 107 copies ) of each plasmid standard . qPCR was performed on a Rotor-Gene Q using the Rotor-Gene SYBR Green PCR Kit with 1 µM primers and 1 µl of undiluted cDNA per 10 µl reaction ( QIAGEN ) . After 5 minutes of denaturation at 95°C , a two-step amplification cycle with 95°C for 5 sec denaturation and 60°C for 20 sec of annealing and extension was used for 45 cycles . Melting curve analyses were performed to ensure that amplified products were specific for the gene of interest . At least three independently acquired biological replicates were analyzed per stage for each gene , with each sample internally replicated 4 times . A polyclonal antibody against Lef-9 was generated by PCR amplifying and cloning a portion of MdBV lef-9 into pET-30-EK-LIC ( Novagen ) as previously outlined [65] . Briefly , 31% of the lef-9 coding sequence was amplified and cloned to create an expression product of 22 . 9 kDa . This construct was confirmed by DNA sequencing , and then expressed in Escherichia coli BL21 ( DE3 ) cells grown in 6 L Luria Broth with 50 µg/ml kanamycin at 37°C . The cultures were then induced with 0 . 025 mM isopropyl-β-d-thiogalactopyranoside ( IPTG ) for an additional 24 h at 37°C . Bacterial cells were harvested by centrifugation , lysed , and the insoluble recombinant protein was purified from the cell pellet using PerfectPro Ni-NTA ( QIAGEN ) agarose beads under denaturing conditions . After analysis by SDS–PAGE and immunoblotting using an anti-His monoclonal antibody , the identity of the recombinant protein was validated by mass spectrometry . A polyclonal antibody was then produced by a commercial service ( Pacific Immunology ) which generated antisera in rabbits by immunizing with ca . 500 µg of affinity purified truncated LEF-9 . Antiserum was then purified by nitrocellulose-based immunoaffinity chromatography as previously outlined [65] , [66] . The resulting antibody was then used in immunoblotting experiments with control and lef-9 knockdown wasps by explanting ovaries , separating ovary extracts on 5–20% SDS-PAGE gradient gels and transferring to PVDF ( Immobilon ) . LEF-9 was then visualized using a goat anti-mouse secondary antibody and the ECL Advance kit as previously described [65] , [67] . The titer of MdBV virions with DNAse-protected episomes of MdBV segment B was quantified by qPCR . One half of an ovary pair for each wasp individual was homogenized with a pestle in 100 µl of DNase buffer from the Roche HighPure RNA Isolation Kit , and NP40 was added to a final concentration of 1% to solubilize wasp cells and virus particle envelopes . After 20 min of gentle rocking , 1 µl of TURBO DNase from the Ambion DNA free kit was added and samples were incubated at 37°C for 40 min to digest all free wasp and episomal viral DNAs . After the addition of EDTA ( 10 mM ) to inactivate the DNase , 250 µg of proteinase K ( Roche ) and 2% sarcosyl were added to samples , followed by incubation at 62°C for 1 h and by phenol∶chloroform extraction and ethanol precipitation in the presence of 0 . 3 M sodium acetate , pH 5 . 2 . DNA pellets were resuspended in 30 µl of 10 mM Tris-Cl pH 8 . 5 and diluted to 1 ng/µl for use as template . Segment B specific primers flanking the point of circularization in the viral segment were used to amplify circularized viral DNA using qPCR as described above and previously [7] ( Figure 7B , Table S2 ) . Per-ovary copy numbers were calculated by multiplying the qPCR estimate of copy number for a half ovary by two , and by the dilution factor and elution volume . At least 3 independently acquired biological replicates were performed for each treatment with samples internally replicated 4 times . Quantification of viral segment excision was performed on genomic DNAs extracted from the entire tissue from an ovary half with the proteinase K , sarcosyl , and phenol/chloroform extraction method described above , without a DNAse step . Primers specific for the empty locus for Segment B were used in qPCR to quantify copy number ( Figure 7B , Table S2 ) . TEM was performed as in [31] . The CiE1 cell line was maintained as in [43] . The infectivity of virus preparations was measured by counting the number of Segment B viral genome copies in CiE1 cells after 24 hours of incubation [43] . Virus was collected from 24 h old wasps treated dsRNA by puncturing a half ovary in 50 µl of PBS and allowing the virus to dissolve . The entire contents of the droplet were added to a microcentrifuge tube containing 200 ul of Sf900 media with 5% fetal bovine serum and 1% antibiotics . This solution was filtered through a 0 . 45 µm membrane to remove any bacteria and cellular debris . Fifty µl of each virus preparation , corresponding to 0 . 1 wasp equivalents or an estimated MOI of 100 for control samples [29] was added to wells in a 24-well cell culture plate containing 105 CiE1 cells per well . Virus particles were incubated with cells for 2 h at room temperature , followed by removal of virus-containing media and the addition of 500 µl of fresh media . CiE1 cells were incubated for an additional 22 hours at 26°C . DNA was isolated from cells following the protocol for quantification of viral titer above without prior DNase treatment . All samples were diluted to a concentration of 50 ng/µl for qPCR amplification of circularized Segment B as described above . Successful viral gene expression , translation and export of protein products in host cells was quantified by counting the percentage of cells displaying the MdBV protein GLC1 . 8 on their surface . CiE1 cells were infected with virus preparations as described above , and after 24 hours of total incubation were fixed with 3 . 7% paraformaldehyde and stained with a murine monoclonal antibody specific for GLC1 . 8 followed by goat anti-mouse Alexa-fluor 568 secondary antibody ( Molecular Probes ) as described [44] . One hundred CiE1 cells were counted from a randomly selected field of view using an epifluorescent , phase-contrast microscope ( Leica DM IRB ) . JMP v10 was used for all statistical analyses . For qPCR analyses , the number of copies of a gene or DNA product were averaged for all technical replicates within a biological replicate . For functional assays , means were calculated from experimental values derived from biological replicates . Each biological replicate represented an individual wasps' ovary . Differences between means of biological replicates were tested using a t-test assuming equal variances or ANOVA . Differences between means for experiments with more than two treatments were distinguished using Tukey's HSD test at the p<0 . 05 significance level . | Microorganisms form symbiotic associations with animals and plants that range from parasitic ( pathogens ) to beneficial ( mutualists ) . Although numerous examples of obligate , mutualistic bacteria , fungi , and protozoans exist , viruses are almost always considered to be pathogens . An exception is the family Polydnaviridae , which consists of large DNA viruses that are obligate mutualists of insects called parasitoid wasps . Prior studies show that polydnaviruses in the genus Bracovirus evolved approximately 100 million years ago from a group of viruses called nudiviruses , which are themselves closely related to a large family of insect pathogens called baculoviruses . Polydnaviruses are thus of fundamental interest for understanding the processes by which viruses can evolve into mutualists . In this study we characterized the composition of virus particles from Microplitis demolitor bracovirus ( MdBV ) and conducted functional experiments to assess whether BV genes share similar functions with related essential baculovirus replication genes . Our results indicate that several genes in MdBV retain ancestral functions , but select other genes have novel functions unknown from baculoviruses . Our results also provide the first experimental data on the function of polydnavirus replication genes and enhance understanding of the similarities between these viruses and their pathogenic ancestors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"dna",
"viruses",
"viral",
"classification",
"virology",
"biology",
"microbiology"
] | 2013 | Mutualistic Polydnaviruses Share Essential Replication Gene Functions with Pathogenic Ancestors |
ATR activation is dependent on temporal and spatial interactions with partner proteins . In the budding yeast model , three proteins – Dpb11TopBP1 , Ddc1Rad9 and Dna2 - all interact with and activate Mec1ATR . Each contains an ATR activation domain ( ADD ) that interacts directly with the Mec1ATR:Ddc2ATRIP complex . Any of the Dpb11TopBP1 , Ddc1Rad9 or Dna2 ADDs is sufficient to activate Mec1ATR in vitro . All three can also independently activate Mec1ATR in vivo: the checkpoint is lost only when all three AADs are absent . In metazoans , only TopBP1 has been identified as a direct ATR activator . Depletion-replacement approaches suggest the TopBP1-AAD is both sufficient and necessary for ATR activation . The physiological function of the TopBP1 AAD is , however , unknown . We created a knock-in point mutation ( W1147R ) that ablates mouse TopBP1-AAD function . TopBP1-W1147R is early embryonic lethal . To analyse TopBP1-W1147R cellular function in vivo , we silenced the wild type TopBP1 allele in heterozygous MEFs . AAD inactivation impaired cell proliferation , promoted premature senescence and compromised Chk1 signalling following UV irradiation . We also show enforced TopBP1 dimerization promotes ATR-dependent Chk1 phosphorylation . Our data suggest that , unlike the yeast models , the TopBP1-AAD is the major activator of ATR , sustaining cell proliferation and embryonic development .
In response to endogenous and exogenous stress , cells have evolved a range DNA damage response ( DDR ) pathways to maintain genomic stability [1] , [2] , [3] , [4] . In all eukaryotes , two evolutionarily conserved PI3-kinase-like protein kinases , ATM ( Ataxia telangiectasia mutated ) and ATR ( ATM- and Rad3-related ) respond directly to DNA damage to control cell-cycle progression and regulate other DNA damage responses such as DNA repair and apoptosis . ATM activation is triggered by double-strand breaks ( DSBs ) , whereas ATR activation is induced by single stranded DNA ( ssDNA ) occurring due to replication stress , resected-DSBs or other single strand lesions [5] , [6] , [7] , [8] . In all eukaryotic organisms , ATR is found associated with ATRIP ( ATR-interacting protein ) which is necessary to recruit ATR to RPA-coated ssDNA [9] , [10] , [11] , [12] . In metazoans , but not in yeasts , this correlates with ATR autophosphorylation at T1989 [13] , [14] . Pre-requisite for ATR activation is the independent recruitment of the Rad17-RFC checkpoint clamp loader to the junction of RPA-coated ssDNA and double stranded DNA ( dsDNA ) , where it facilitates the loading of the Rad9-Hus1-Rad1 ( 9-1-1 ) sliding clamp [15] . The co-recruitment of ATR and the 9-1-1 clamp establishes a platform for activation of the ATR pathway [13] , [16] , [17] . The C-terminus of the Rad9 subunit of the 9-1-1 clamp is responsible for recruiting TopBP1 [18] , [19] , [20] , a conserved multi-BRCT-domain scaffolding protein [21] . In yeast model systems , the Rad9 C-terminus must be phosphorylated by ATR to provide a docking site for phospho-protein binding domains within TopBP1 [18] , [20] . In metazoans , Rad9 is constitutively phosphorylated by CK2 and thus TopBP1 recruitment does not require ATR-dependent Rad9 phosphorylation . Metazoan TopBP1 contains nine BRCA1 C-terminal ( BRCT ) domains while the yeast homologs contain only four BRCT domains . The TopBP1 BRCT domains define phospho-binding motifs [22] , [23] that allow TopBP1 to scaffold distinct proteins and protein complexes in response to the phosphorylation status of its clients . In all eukaryotes , TopBP1 plays an essential role in the initiation of DNA replication [21] . In yeast models , this function reacts to cell cycle-dependent phosphorylation of two client proteins , Sld2 and Sld3 , allowing TopBP1 to bridge an interaction between two replication factors in order to promote Cdc45 and GINS loading to activate the replicative helicase [24] . A similar role is evident in metazoans , where TopBP1 association with the Sld2 homolog , Treslin , is essential for replication initiation [25] , [26] , [27] . In response to ssDNA during DNA replication stress or DNA repair , yeast TopBP1 performs an equivalent scaffolding role , bridging between the 9-1-1 checkpoint clamp and the checkpoint mediator proteins ( the 53BP1 homologs ) which present checkpoint effector kinases ( Chk's ) to ATR [28] , [29] , [30] . This scaffolding role is essential for a functional ATR-Chk response . In addition , yeast TopBP1 contains a conserved ATR activation domain ( AAD ) which , when over-expressed , can directly induce ATR activation in the absence of DNA damage 11 , 31 , 32 . The yeast TopBP1 AAD participates in , but is not necessary for , ATR activation: AAD-deficient separation of function mutants display only sensitivity to genotoxins and partial , cell cycle-specific checkpoint defects [32] , [33] . As observed in the yeast models , metazoan TopBP1 is similarly required for activation of the ATR-Chk1 axis and is recruited to the site of DNA damage by binding to the C-terminus of the 9-1-1 complex [16] , [31] , [34] . However , at this point , significant differences emerge between the yeast and metazoan systems: in addition to the constitutive formation of a 9-1-1 TopBP1 interaction in metazoans ( see above ) , replacing Xenopus TopBP1 with a recombinant protein containing a mutation in the AAD ( W1138 ) completely blocks ATR activation in response to replication inhibition in extracts [31] . While extracts may not fully recapitulate all aspects of the cellular environment , this suggests a more important role for the metazoan TopBP1 AAD in ATR activation when compared to yeast . The differences between ATR activation in the yeast and the metazoan systems are intriguing . In the yeast models , ATR provides the bulk of the checkpoint signalling following all forms of DNA damage , including DSBs . In metazoans , ATM provides the majority of checkpoint signalling in response to DSBs , with ATR playing a minor role . This distinction between yeasts and metazoans can be explained , at least in part , by different repair priorities: yeasts generally rapidly resect DSBs for repair by homologous recombination , with non-homologous end joining ( NHEJ ) - which occurs without significant resection - playing a minor role . Conversely , metazoan cells rely largely on NHEJ and thus detect DSBs through the ATM pathway . Experimentally limiting resection in yeast models uncovers an ATM-dependent checkpoint [35] , [36] , demonstrating the underlying machinery is conserved . The distinct repair priorities between yeast and metazoan systems are likely to underpin changes in the architecture of ATR activation mechanisms during evolution . For example , it is notable that distinct pairs of BRCT-domains mediate the 9-1-1 interaction in yeasts and metazoans: 9-1-1 interacts with BRCT 3+4 , in yeasts ( homologous to metazoan BRCT 4+5 ) but with the conserved BRCT1+2 pair in metazoans [21] . Furthermore , BRCT domains 7/8 in metazoans ( which is not conserved in yeasts ) binds autophosphorylated ATR-T1989 to promote a tight complex and strengthen ATR signalling [13] , [14] . Complete deletion of TopBP1 in untransformed mouse or human primary cells induces cellular apoptosis and TopBP1 deficiency results in an early embryonic lethality [37] , [38] , [39] , [40] , [41] . Tissue specific deletion of TopBP1 in the central nervous system ( CNS ) similarly leads to an accumulation of DNA breaks in neuronal progenitors and subsequent disruption of neurogenesis [42] . These data are consistent with the essential role for TopBP1 in the initiation of DNA replication . To specifically establish the physiological function of the TopBP1 AAD , and to investigate if it is dispensable for ATR activation in metazoans as it is in yeasts , we generated a mouse model with a specific knock-in AAD mutation . We show that the TopBP1 AAD is essential for the embryonic development , phenocopies the lethal phenotype of ATR and is necessary for ATR signalling after UV damage . These data strongly suggest that , unlike in the yeast models , ATR activation of by the TopBP1 AAD is the major , if not only , route to activating the ATR-Chk1 axis and is essential for cell proliferation and survival .
Heterozygous TopBP1ki/+ mice are viable and phenotypically normal during a 2 year-observation period ( data not shown ) . However , no homozygous TopBP1ki/ki offspring were obtained from TopBP1ki/+ intercrosses ( 167 live births genotyped , Table 1 ) . Backcrossing TopBP1ki/+ with TopBP1+/+ gave the expected ratio of TopBP1ki/+ and TopBP1+/+ offspring , indicating that there were no fertility defects in either the male or female TopBP1ki/+ animals ( Table 1 ) . We thus analyzed deciduas and embryos derived from TopBP1ki/+ intercrosses: genotyping at E11 . 5 revealed no homozygous mutants ( Table 1 ) , although 17/48 deciduas were small , precluding reliable PCR due to the presence of mother-derived tissues ( Fig . 2A ) . We next isolated blastocysts ( E3 . 5 ) from TopBP1ki/+ intercrosses for PCR genotyping: 21 . 7% , close to the expected Mendelian ratio , of embryos were TopBP1ki/ki homozygotes ( Table 1 ) . Although morphology was normal at isolation , all TopBP1ki/ki blastocysts failed to outgrow in vitro cultures ( Fig . 2B , C ) . These data indicate that the TopBP1 AAD function is required for embryo development beyond the blastocyst stage . The lethal phenotype of TopBP1-W1147R precluded the use of TopBP1ki/ki MEFs for direct cellular assays . However , the presence of the base change associated with the knock-in mutation ( T3439C ) plus a second nucleotide polymorphism ( Fig . S2A ) derived from the targeting construct ( T3477C , a silent mutation ) offered the opportunity to specifically silence the TopBP1+ allele in TopBP1ki/+ MEFs . We designed two independent shRNA expression vectors targeted against the wild type sequence , but predicted to leave the knock-in allele resistant to RNA interference . Using co-transfection with corresponding chimeric GFP-encoding reporters , these were tested individually in COS7 cells for their ability to specifically target transcripts with the wild type ( GFP-wtAAD ) , but not the knock-in ( GFP-mutAAD ) TopBP1 sequences ( Fig . S3 ) . The shRNA construct targeting the T3477C point mutation , designated shTop2 , efficiently targeted GFP-wtAAD but not GFP-mutAAD . We thus transferred the shRNA construct into a vector expressing GFP to create GFP-shTop2 . The co-expression of GFP from the shRNA vector will allow selection for transfected cells . GFP-shTop2 , or a GFP-shLuciferase ( GFP-shLuc ) control , was transfected into TopBP1ki/+ MEFs previously immortalized by a 3T3 protocol . GFP-positive cells were sorted by FACS 36 hours after transfection ( Fig . S3D ) . Semi-quantitative reverse transcription PCR ( RT-PCR ) indicated that , in comparison to GFP-shLuc control transfected cells , TopBP1 mRNA levels were reduced upon GFP-shTop2 transfection . Sequencing revealed that the remaining TopBP1 mRNA was predominantly the AAD mutant form ( Fig . 3A ) . Western blot analysis revealed a reduction of ∼50% for the TopBP1 protein following GFP-shTop2 transfection of TopBP1ki/+ cells when compared to GFP-shLuc control transfected cells ( Fig . 3B , C ) . Corroboration of the specificity of GFP-shTop2 to the wild type TopBP1 transcript comes from the observation that GFP-shTop2 efficiently knocked down TopBP1 levels in TopBP1+/+ MEF cells ( Fig . 3B , C ) . In the following experiments , we designated GFP-shTop2-transfected TopBP1ki/+ cells as GFP-TopBP1ki/− and GFP-shLuc transfected TopBP1ki/+ cells as GFP-TopBP1ki/+ . To identify the consequences of specific loss of TopBP1 AAD function in cell survival and proliferation we followed the proportion of the GFP positive cells after GFP-shTop2 or GFP-shLuc transfection . 24 hours after transfection , the GFP+ population was similar for GFP-TopBP1ki/− ( GFP-shTop2 transfected ) and control GFP-TopBP1ki/+ cells ( GFP-shLuc transfected ) . However , 5 days after transfection the proportion of GFP-TopBP1ki/− cells was significantly reduced when compared to GFP-TopBP1ki/+ cells ( Fig . 3D ) . Immunostaining against cleaved Caspase 3 revealed a mild ( but not statistically significant ) increase of apoptosis in GFP-TopBP1ki/− when compared to control GFP-TopBP1ki/+ cells 4 days after of transfection ( Fig . 3E ) . This suggests that delayed proliferation as opposed to apoptosis is the major cause of the reduced number of GFP-TopBP1ki/− cells . To further characterise the effect of the AAD mutation on cell proliferation , cells were pulse-labeled with EdU for 2 hours either 36 or 84 hours following transfection . Consistent with reduced proliferation , the percentage of GFP+ cells that were also positive for EdU in GFP-TopBP1ki/− cultures was reduced when compared to GFP-TopBP1ki/+ controls ( Fig . 3F , G ) . Nonetheless , a significant proportion of GFP-TopBP1ki/− cells were also EdU+ , consistent with the AAD mutation being dispensable for DNA replication . Following transfection with GFP-shLuc , control ( GFP-TopBP1ki/+ ) cells became confluent after 5 days in culture . However , the density of GFP-TopBP1ki/− cells following transfection with GFP-shTop2 did not significantly increase over the same period ( Fig . 4A ) . GFP-TopBP1ki/− cells became giant and flat after 4 days in culture ( Fig . 4A ) and this was often associated with β-galactosidase positive staining , indicative of cellular senescence ( Fig . 4B , C ) . Consistent with this , RT-PCR analysis revealed up-regulated expression of p19ARF and p21 , known senescence markers , in senescent GFP-TopBP1ki/− cells ( Fig . 4D ) . We conclude that loss of TopBP1 AAD function results in proliferation defects and promotes entry into senescence . To test if the TopBP1 AAD mutation affected activation of the ATR pathway following DNA damage treatment we analyzed Chk1 phosphorylation following UV irradiation . First we established that , in our assay , UV irradiation resulted in RPA foci formation - an event occurring upstream from , and independent of , ATR activation . As expected , UV treatment resulted in similar patterns of RPA foci in GFP-TopBP1ki/− , GFP-TopBP1ki/+ and non-transfected control cells . Inhibition of ATR activity using the chemical inhibitor ( ATRi; ETP-46464 [43] ) similarly did not affect RPA localization after UV ( Fig . 5A ) . Conversely , Immunostaining for phosphorylated Chk1 ( p-Chk1-S317 ) to detect substrates downstream of ATR activation showed a dramatic increase of pChk1 in untransfected and GFP-TopBP1ki/+ ( shLuc transfected ) cells , whereas GFP-TopBP1ki/− ( shTop2 transfected ) and ATRi-treated TopBP1ki/+ cells showed attenuated p-Chk1-S317 staining to similar levels ( Fig . 5B , C ) . Verifying the specificity of the assay , UV-induced p-Chk1-S317 staining was fully abrogated in TopBP1ki/+ cells following GFP-shChk1 transfection ( Fig . 5B ) [44] . These data are consistent with an expectation that TopBP1 AAD function is necessary to activate ATR in response to UV treatment . It has previously been established that phosphorylation of human TopBP1 within the AAD at S1159 ( analogous to mouse S1161 ) by Akt/PKB facilitates TopBP1 oligomerisation [45] , [46] . To establish if our AAD mutation compromises ATR activation by preventing TopBP1 oligomerisation , we adopted an experimental approach that exploits inducible dimerization: TopBP1 was fused to FKBP-F36V ( Fig . 6A ) , a mutant form of FKBP12 that forms a dimer upon binding to the synthetic ligand AP20187 [47] . Flag- and HA-tagged wild type TopBP1 ( wtTopBP1 ) or TopBP1-W1147R ( mutTopBP1 ) , each fused with FKBP , were co-expressed in all combinations . Cells were then treated with AP20187 and extracts assayed for expression and co-precipitation of HA-tagged protein by the Flag-tagged protein . As expected , interactions mediated by AP2187 ligand were observed for all combinations ( Fig . 6B ) consistent with the expectation that a point mutation in the AAD does not disrupt FKBP-induced dimer formation . Interestingly , when we transfected HA-FKBP-wtTopBP1 into cells we observed that the dimerization of TopBP1 induced by AP20187 promoted ATR-dependent Chk1 phosphorylation ( ATRi treatment abolished phosphorylation; see lane 6 ) . Unexpectedly , this was independent of DNA damage treatment ( Fig . 6C: compare lanes 2 with 4 ) . We next exploited this dimerisation-induced Chk1 phosphorylation to establish if the requirement for the TopBP1 AAD in ATR activation could be bypassed by forced dimerisation . HA-FKBP-wtTopBP1 , HA-FKBP-mutTopBP1 and the controls HA-FKBP and HA-wtTopBP1 were each transfected into cells and AP20187 ligand-dependent Chk1 phosphorylation monitored . Neither FKBP alone or TopBP1 alone resulted in Chk1 phosphorylation in response to AP210187 ligand . As expected , HA-FKBP-wtTopBP1 expression resulted in Chk1 phosphorylation upon ligand addition ( Fig . 6D ) . Conversely , HA-FKBP-mutTopBP1experssion did not induce Chk1 phosphorylation in response to ligand . Suggestive of a dominant negative effect , Chk1 phosphorylation in these cells was impaired below background upon AP20187-induced interaction ( Fig . 6B , D ) . These results indicate that induced oligomerisation of TopBP1 is sufficient to induce ATR activation and subsequent Chk1 phosphorylation and that this requires the TopBP1 AAD function .
In the present study , we investigated the biological significance of the TopBP1 AAD in a mouse model . We created a single point mutation ( W1147R ) within the AAD of TopBP1 that removes a key aromatic residue necessary for the activation of ATR . This is predicted to separate the AAD function from other essential functions such as the scaffolding role during replication initiation and from any roles in scaffolding checkpoint complexes . Unexpectedly , this point mutation resulted in early embryonic lethality and developmental arrest at the blastocyst stage . This early lethal phenotype is equivalent to that reported for the complete knockout of TopBP1 in mice [51] and is reminiscent of the consequence of ATR deletion [52] , [53] . Due to the early lethality we could not directly eliminate the possibility that the homozygous AAD knock-in ( TopBP1ki/ki ) mutation had , in fact , generated a null mutation . However , both the mRNA and TopBP1 protein levels were produced at the expected levels by the AAD knock-in allele which was visualized by specific shRNA knock-down of the wild type TopBP1 mRNA in heterozygous MEFs ( TopBP1ki/+ ) ( see Fig . 3B and Fig . S3B ) . Based on these data we propose that the TopBP1-W1147R ( AAD mutant ) protein is stable and the phenotypes observed are a direct consequence of the mutation introduced . Our results thus suggest that one essential function for TopBP1 in embryonic development is realized by a TopBP1 AAD-mediated ATR activation function and that this cannot be substituted for by other potential AAD domains . In addition , the scaffolding functions of TopBP1 in replication initiation and checkpoint activation cannot sustain embryonic development and are insufficient for ATR activation . By establishing an shRNA knock-down assay which specifically targeted the wild type , but not the TopBP1-W1147R ( AAD mutant ) mRNA , we were able to examine the effect of the AAD mutation in MEFs . Our first observation is that MEFs containing only mutated TopBP1-W1147R ( GFP-TopBP1ki/− ) were not able to proliferate and entered senescence . This is consistent with the early embryonic lethality and strongly suggests an essential cellular role for the TopBP1 AAD , presumably by activating ATR . Our preferred explanation is that specific lesions are generated in mammalian cells during DNA replication and that , in response to these , only the TopBP1 AAD is capable of activating ATR . Such an explanation does not preclude the existence of additional ATR activating domains in other proteins ( as is observed in the yeasts ) but would suggest that , if these exist , they respond to alternative DNA structures or to structures formed at different points in the cell cycle , for example only in G1 . Induced DNA damage , such as that caused by UV irradiation , arrests cell proliferation via cell cycle checkpoint activation . We examined the response of cells to UV irradiation and observed that , in the absence of the wild type protein ( via shRNA knock-down ) , cells expressing the TopBP1-W1147R ( AAD mutant ) protein were unable to mount a significant ATR response . The parsimonious explanation for this is that , in mammalian cells , the TopBP1 AAD is either the main or the sole mechanism for activating ATR . Given the additional complexity evident in the yeasts , this is surprising to us: evolution is prone to elaborate mechanistic pathways as organisms become multicellular and more complex . Nonetheless , our data suggest that the TopBP1 AAD is responsible for the majority of ATR signaling and that additional ATR activating domains play little or no role in metazoan checkpoint responses . The inhibition of TopBP1 expression by antisense oligomers or by siRNA induces apoptosis in cancer cell lines or MEF cells [37] , [38] , [39] , [40] , [41] . In contrast , we did not observe a statistically significant increase in apoptosis when cells grew in the presence of the TopBP1 AAD defective protein ( GFP-TopBP1ki/− ) . Instead , we observed increased cellular senescence that was associated with elevated expression of p19 and p21 . Full loss of TopBP1 function would be expected to disrupt replication initiation , whereas the specific loss of the AAD function may allow replication but lead to an accumulation of spontaneous damage that subsequently signals through the ATM pathway . Consistent with this , we did observe some incorporation of EdU in GFP-TopBP1ki/− cells and we thus suggest that the reduced proliferation and increased cellular senescence observed in GFP-TopBP1ki/− cells stems from impaired G1/S transition likely resulting from ATM activation . MEF cells deleted for ATR similarly show an increase in cellular senescence , a reduction of proliferation and only a small increasing of in apoptosis [54] . This is also consistent with our expectation that the TopBP1 AAD mutation specifically affects the ATR-Chk1 cascade without preventing replication initiation . While establishing that the AAD mutation in TopBP1 was not preventing ATR activation due to a dimerization defect , we found that forced dimerization of TopBP1 strongly stimulated ATR activation in the absence of induced DNA damage , as judged by a significant increase in Chk1phosphorylation . While oligomerisation can lead to increased protein stability and improvements to enzymatic activity [55] , we did no observe any increase of the TopBP1 protein level following induced oligomerisation . Several alternative possibilities could account for ATR activation by oligomerised TopBP1: oligomerisation may enhance the affinity of TopBP1 for its interaction partners . In this regard , it is interesting to note that phosphorylation of Ser1131 ( ortholog of human TopBP1 Ser1140 ) in the AAD of Xenopus TopBP1 enhances binding of to the Xenopus ATR-ATRIP complex , and thereby increases the capacity of TopBP1 to activate the ATR [56] . Alternatively , oligomerisation of TopBP1 may enhances its chromatin binding ability . In this regard it is interesting to note that tethering TopBP1 [57] or the S . pombe homolog ( Rad4TopBP1 ) to chromatin [32] activates the ATR and Chk1-dependent checkpoint . As expected , despite the forced oligomerisation of the AAD mutant of TopBP1 , it failed to stimulates ATR activity , strongly suggesting that TopBP1 oligomerisation is necessary but not sufficient for ATR activation and that an intact AAD is required .
Gene targeting vector was constructed with Red/ET recombineering technology ( Gene Bridges ) . Briefly , a LoxP-Neo-LoxP cassette was inserted into bacmid ( bMQ-304N19 , Geneservice ) encompassing the genomic region of TopBP1 , using the Red/ET Quick and Easy BAC Modification Kit . The Neo cassette was subsequently excised by expression Cre recombinase in host bacterial cells , resulting in a one LoxP site in intron 20 . Next , a second Flp-Neo-Flp cassette was inserted into intron 19 . Mutation of Tryptophan to Arginine at 1147 was achieved by in vivo substitution of T3439 by C3439 ( Counter-Selection BAC Modification Kit , Gene Bridges ) . The engineered genomic region of TopBP1 in bacmid was then subcloned into high-copy plasmid vector ( ColE1 ) by homologous recombination , resulted in the targeting construct of TopBP1-W1147R . Flag-tagged full length wild type or W1147R mutant TopBP1 were amplified by PCR with 5′-primer TopFL-5 ( TACGGATCCCTCGGGCTCCACCTAGTTCA ) and 3′- primer TopFL-3 ( CCGCTCGAGGCCGTTTGACTACATTC ) and constructed into pCMV-tag 2C ( Stratagene ) , pcDNAHA , or pcDNAHA2FKBP vector , respectively [47] . GFP-tagged wild type or W1147R mutant AAD of TopBP1 were amplified by PCR with 5′-primer micTop54-2 ( GAAGATCTTGACCCAGGCCTTGGAGATGAGAG ) and 3′-primer micTop34-2 ( ACGCGTCGACTGCCCTGGGGCTTGAGTAACACA ) and constructed into pEGFP C2 ( Clontech , Mountain View , CA , USA ) . The construction of shRNA expression vectors was performed as previously described [58] . Briefly , oligonucleotides targeting the coding sequences and their complementary sequences were inserted into the vector under the control of the human U6 promoter with or without CMV-driven EGFP . All the oligonucleotides contained the following hairpin loop sequence: TTCAAGAGA . The targeting sequences used were: Luciferase: GGCTTGCCAGCAACTTACA , shTop1: TGAGCAGATCATTTGGGACG , and shTop2: TGGCTTGCCAGCAACTTACA . All the constructions were confirmed by sequencing . shRNA expression vector to target Chk1 was reported as previously [44] . The gene targeting vector was linearized by Cla I digestion and electroporated into the E14 . 1 ES cells . After selection with G418 , correctly targeted TopBP1W1147R knock-in ( ki ) ES clones were identified by Southern blot analysis and used to generate germline chimeric mice . To analyze the 5′-arm integration of the targeting vector into the TopBP1 locus , ES cell DNA was digested with AseI and probed with an intron-16 probe ( p8 ) located externally to the upstream of targeting area . 3′-arm integration of the targeting vector was analyzed by digestion the DNA with PpuM 1 and hybridization with an intron-27 probe ( p5 ) located externally to the downstream of targeting area ( see Fig . 1A ) . For the PCR genotyping , the following primers were used: Top158: CTTCTCACTGTGCTGCTTCCTATAGC; Top159: GCTATTAATTGAGTTTTGTGAATCCC; In19-1f: GCAAGCCATGCAAGTCAATA; In19-2r: GCTTCCCCTGCTGTGATA; neo-1f: ATCTCCTGTCATCTCACCTTGC . The primer pair Top158 and Top159 was used to detect the wild type allele ( wt ) and targeted allele ( tg ) or knock-in targeted allele ( ki ) . Combination of In19-1f , neo-1f and In19-2r detects the remove of neo-cassette in targeted allele . For sequencing genotyping of the TopBP1W1147R ki allele , genomic DNAs were isolation and sequenced with primer In19-1f . The total RNA was isolated by using Tri Reagent ( T9424 , Sigma-Aldrich , Munich , Germany ) . 1 µg of RNA was used for synthesis of first-strand cDNA by Affinity Script Multiple Temperature cDNA Synthesis Kit ( 200436 , Stratagene ) according to the manual . Semi-quantitative PCR was performed with the following primers . For TopBP1: micTop54-2 and micTop34-2 ( see 4 . 1 ) ; for GAPDH: forward primer mGAPDH15 ( GCACAGTCAAGGCCGAGAAT ) and reverse primer mGAPDH13 ( GCCTTCTCCATGGTGGTGAA ) ; For p19ARF: forward primer p19f ( CCCACTCCAAGAGAGGGTTT ) and reverse primer p19r ( TCTGCACCGTAGTTGAGCAG ) ; For p21: forward primer p21f ( GTCAGGCTGGTCTGCCTCCG ) and reverse primer p21r ( CGGTCCCGTGGACAGTGAGCAG ) . Primary mouse embryonic fibroblasts ( MEFs ) were isolated from E13 . 5 embryos derived from the mating between TopBP1ki/+ mice and immortalized with a standard 3T3 protocol [59] . For transfection , 3T3 MEFs were transfected using Amaxa Nucleofector Kit R ( VCA-1001 , LONZA , Cologne , Germany ) . Briefly , MEFs were trypsinized and 1×106 cells were centrifuged at 200× g for 10 min . The cell pellet was resuspended in 100 µl Nucleofector Solution mixture plus 5 g of plasmid-DNA . The cell suspension was electroporated using Nucleofector I Device ( Lonza ) . The electroporated MEFs were cultured under normal conditions for 24 hr before FACsorting based on GFP expression . The sorted cells were either used for protein extraction , mRNA isolation or further cultured in the presence of 400 ug/ml of G418 ( Invitrogen ) . For EdU labeling , cells were incubated with 1 µg/ml of EdU ( A10044 , Invitrogen ) for 2 hr at 36 or 84 hr after transfection . UV exposure and HU treatment were performed at 36 hr after transfection and cells were fixed with 4% PFA for immunofluorescence staining . For ATR inhibitor treatment , 1 . 6 µm ATR inhibitor ( ATRi ) was added 1 hr before exposure to 100 J/m2 of UV . Cos7 or HEK293T cells were transfected with lipofectamine2000 ( 11668-019 , Invitrogen ) according the manufacturer's instruction Immunostaining was performed on as described previously [44] . Briefly , PFA-fixed cells were incubated with blocking buffer ( 1% BSA , 5% goat serum and 0 . 4% Triton X-100 in PBS ) for 1 hr at room temperature then with a primary antibody diluted in blocking buffer at 4°C overnight followed by secondary antibodies for 2 hr at room temperature . After washing , the slides were mounted with DAPI-containing mounting medium ( Invitrogen ) . The primary antibodies and respective dilutions are: rabbit anti-pChk1-S317 antibody ( 1∶100 , A300-163A-3 , Bethyl Laboratories , Montgomery , TX , USA ) ; rabbit anti-Cleaved Caspase-3 ( Asp175 ) ( 1∶300 , 9662 , Cell Signaling Technology , Danvers MA , USA ) and rat anti-RPA antibody ( 1∶300 , 2208 , Cell Signaling Technology ) . EdU detection was carried out using a Click-iT EdU Alexa Fluor 647 Imaging Kit ( 953624 , Invitrogen ) after fixation according to the manufacturer's instruction . β-galactosidase staining was performed with a Senescence β-galactosidase staining kit ( 9860 , Cell Signaling Technology ) according to the manufacturer's instruction . Cells images were acquired using a virtual microscope ( BX61VS , Olympus , Tokyo , Japan ) or a confocal microscope ( LSM510 , Zeiss , Jena , Germany ) . The density of fluorescent signal was quantified by a high-content analysis microscopy ( Cellomics Arrayscan VTI , Pittsburgh , PA , USA ) . The proteins were extracted with RIPA buffer ( 20 mM HEPES , pH 7 . 6 , 20% glycerol , 0 . 5M NaCl , 1 . 5 mM MgCl2 , 0 . 2 mM EDTA , pH 8 . 0 , 0 . 5% NP-40 , 1 mM DTT , 1 mM PMSF , 5 mg/ml leupeptin , 2 mg/ml aprotinin , 1 mM β-glycerophosphate , 1 mM Na3VO4 and 10 mM NaF ) from cells . After separation in SDS-PAGE , the membranes were blotted with the flowing antibodies . The primary antibodies used in this study were rabbit anti-TopBP1 antibody ( 1∶1000 , AB3245 , Millipore , Schwalbach , Germany ) , rabbit anti-phospho-S317-Chk1 ( 1∶1000 , A300-163A , Bethyl Laboratories ) , Rabbit anti-HA ( 1∶10000 , A190-208A , Bethyl Laboratories ) ; mouse anti-β-Action ( 1∶20000 , C2206 , Sigma-Aldrich ) , mouse anti-Flag ( 1∶10000 , F4042 , Sigma-Aldrich ) , mouse anti- γH2AX ( 1∶1000 , 05-636 , Millipore ) ; sheep anti-Chk1 antibody ( 1∶1000 , ab16130 , Abcam , Cambridge , UK ) . The inducible dimerization assay was performed as previous described [47] . Briefly , HEK 293T cells were transiently transfected with pcDNAHA2-TopBP1 or pcDNAHA2FKBP-TopBP1 ( or its AAD mutant counterpart ) . Forty hours later , transfectants were either mock-treated with 0 . 1% ethanol or treated with a 100 nM of the bivalent ligand AP20187 ( 635060 , Clontech ) and/or combined with 1 . 6 µm ATR inhibitor ( ATRi ) for 1 hr . Immunoprecipitation was carried out as previous described [47] . Animal experiments conducted in this report were approved and conducted according to the German or British animal welfare legislation and in pathogen-free conditions . | DNA damage checkpoint signalling is an essential component of the DNA damage response . Many of the key proteins initiating the checkpoint signal have been identified and characterised in yeast . Here we explore the role of the ATR activating domain ( AAD ) of TopBP1 in embryonic development , cell growth and checkpoint activation using a mouse model . In contrast to yeasts , where the TopBP1 AAD plays a redundant , and thus phenotypically minor , role in ATR activation , our data demonstrate that the mouse TopBP1 AAD is essential for cellular proliferation . Interestingly , this suggests evolution has provided a simpler ATR activation mechanism in metazoans than it has in yeasts . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2013 | An Essential Function for the ATR-Activation-Domain (AAD) of TopBP1 in Mouse Development and Cellular Senescence |
The collagen binding integrin α2β1 plays a crucial role in hemostasis , fibrosis , and cancer progression amongst others . It is specifically inhibited by rhodocetin ( RC ) , a C-type lectin-related protein ( CLRP ) found in Malayan pit viper ( Calloselasma rhodostoma ) venom . The structure of RC alone reveals a heterotetramer arranged as an αβ and γδ subunit in a cruciform shape . RC specifically binds to the collagen binding A-domain of the integrin α2 subunit , thereby blocking collagen-induced platelet aggregation . However , until now , the molecular basis for this interaction has remained unclear . Here , we present the molecular structure of the RCγδ-α2A complex solved to 3 . 0 Å resolution . Our findings show that RC undergoes a dramatic structural reorganization upon binding to α2β1 integrin . Besides the release of the nonbinding RCαβ tandem , the RCγ subunit interacts with loop 2 of the α2A domain as result of a dramatic conformational change . The RCδ subunit contacts the integrin α2A domain in the “closed” conformation through its helix C . Combined with epitope-mapped antibodies , conformationally locked α2A domain mutants , point mutations within the α2A loop 2 , and chemical modifications of the purified toxin protein , this molecular structure of RCγδ-α2A complex explains the inhibitory mechanism and specificity of RC for α2β1 integrin .
Most cellular processes depend on the formation of interactions between cells and the extracellular matrix ( ECM ) . Key facilitators of these interactions are the integrins . They consist of 2 subunits , α and β , each of which has multiple isoforms [1 , 2] . The different subunit composition between integrins determines their ligand-binding specificity and functionality . Integrins are cell adhesion molecules , which are involved in a broad range of cell functions , such as proliferation , differentiation , adhesion , and migration . Defect or dysfunction of integrins , in particular of α2β1 integrin , a prominent collagen binding receptor of many cell types [3] and the only collagen binding integrin on platelets [4] , may affect vascular development and angiogenesis [5] , epithelial cell differentiation [6] , wound repair and fibrosis [7] , inflammation [8 , 9] , and cancer and cancer therapy [10] , as well as collagen-induced platelet activation , hemostasis , and thrombosis [4 , 11] . Therefore , α2β1 integrin has become a prominent target in drug research [12–14] . The collagen binding site is located within the α2A domain of α2β1 integrin , which is homologous to the A-domain of von Willebrand factor ( vWF ) . The α2A domain contains a metal ion that is required for collagen binding as it is part of the binding site for the collagen triple helix [15] . In order to bind to collagen , the α2A domain undergoes a series of concerted conformational changes . In short , helix C unwinds , the N-termini of helices 6 and 7 simultaneously turn away from each other , and , finally , helix 7 moves downward against helix 1 to give the collagen binding “open” conformation , which contrasts with the previous “closed” conformation [15 , 16] . This likely general mechanism of molecular movement of integrin A-domains was subsequently confirmed by introducing a disulfide bridge into the A-domain of the integrin αL subunit such that this interconversion was blocked with the protein locked in either the “open” or “closed” state [17] . Integrin function can be blocked by two major classes of snake venom proteins , the disintegrins [18 , 19] and the C-type lectin-related proteins ( CLRPs ) [20 , 21] . In contrast to the disintegrins , which can target multiple integrins , CLRPs specifically inhibit α2β1 integrin activity [21] . The high selectivity and affinity of these snake venom proteins for α2β1 integrin make them ideal lead compounds for drug development [22–24] . Current members of the CLRP family include the proteins rhodocetin ( RC ) , EMS16 , vixapatin , sochicetin-B , lebecetin , flavocetin , and rhinocetin [25–31] . As more CLRP structures become available , it is clear that , although the supramolecular structure can vary from the basic heterodimer of EMS16 [27] to the ring-like ( αβ ) 4 structures of flavocetin and convulxin [32 , 33] , the underlying basic unit is a heterodimer consisting of 2 subunits , usually named α and β , which dimerize via their characteristic index finger loops [20 , 34] . Interestingly , in the case of the RC heterotetramer ( αβγδ ) structure [26] , the αβ and γδ subunits form 2 heterodimeric pairs that are oriented orthogonally towards each other in a cruciform shape . Despite these differences , the subunits of CLRP family members are highly homologous with each other . Evolutionarily , the CLRP fold has developed from a carbohydrate recognizing domain ( CRD ) into a structure that specifically targets clotting factors IX and X , α2β1 integrin , and other platelet adhesion receptors [20 , 34–36] . Among the latter , the vWF receptor and the 2 collagen binding receptors , glycoprotein GPIV and α2β1 integrin , are targets for snake venom CLRPs , thereby inhibiting or activating platelet activation and aggregation [37 , 38] . Consequently , these snake venom proteins severely interfere with hemostasis [36 , 39] . However , the nature of the molecular mechanism by which CLRPs inhibit α2β1 integrin and by which CLRPs implement specificity towards α2β1 integrin has remained undetermined . RC is a CLRP of the Malayan pit viper C . rhodostoma [26] , and together with EMS16 from Echis multisquamatus , they are the only known CLRP family members proven to target the α2A domain for which atomic resolution structures are available [27 , 40] . Unlike the α2β1 integrin–collagen interaction , which is metal ion-dependent , the binding of RC to α2β1 integrin does not require a metal ion , which implies a different mechanism of action . In a previous study , we demonstrated that the RCαβγδ heterotetramer binds to α2β1 integrin before releasing the αβ subunit ( RCαβ ) from the complex [40] . In the current work , we present the molecular structure of this RCγδ-α2A domain complex and unravel the molecular mechanism of this interaction . The RC binding site overlaps with that of collagen , including the key metal ion site , thereby sterically blocking collagen binding . Moreover , a comparison with the previously determined RC structure [26] reveals that , in addition to the release of the RCαβ subunit , the RCγδ subunit undergoes a major conformational change upon integrin binding , which causes it to snap into a bent conformation like a mouse trap . In this final state , RCγδ holds the α2A domain in the “closed” conformation , allosterically unable to bind to collagen . The result is a highly efficient inhibition of α2β1 integrin-mediated attachment and signaling in cells and platelets .
To isolate RC in complex with the integrin α2A domain , recombinant α2A domain was immobilized to Ni Sepharose resin via its His6-tag . Thereafter , an RC-rich protein fraction of C . rhodostoma venom was applied to this column , resulting in the formation of the complex of α2A with tetrameric RC ( RCαβγδ ) that still bound to the column . Treatment with 5 mM EGTA resulted in the dissociation of the α2A domain bound RC tetramer and the release of RCαβ from the complex , which was eluted from the column . In contrast , RCγδ remained firmly attached to the column bound α2A ( Fig 1 ) . This RCγδ-α2A complex was then eluted with a linear gradient of imidazole ( Fig 1A ) . Its His6-tag was cleaved by trypsinolysis , and the excess α2A was removed by size-exclusion chromatography . The close physical contact of both partners within the RCγδ-α2A complex was proven by cross-linkage with 0 . 5 mM bis ( sulfosuccinimidyl ) suberate ( BS3 ) ( Fig 1B ) . The crystal structure of the RCγδ-α2A complex was determined at 3 . 0 Å resolution by molecular replacement using the previously determined RCαβγδ structure ( pdb:3GPR ) as a search template ( Fig 2 ) . The RCγδ-α2A structure clearly showed that the RCγδ subunit bound to the top of the α2A domain directly above the metal ion-binding site , thereby sterically blocking access of collagen ( Fig 2A ) . Both chains of RCγδ are typical CLRP folds , characterized by a globular core domain interlinked mutually by extended index finger loops . The A-domain of α2β1 integrin assumed the “closed” conformation with its central β-sheet flanked by the α-helices 3 , 1 , and 7 and 4 , 5 , and 6 on either side . The crystal structures contain 6 RCγδ-α2A complexes per asymmetric unit ( S1 Fig ) . We determined the total interaction surface between RCγδ and α2A in the complex to be 965 Å2 . There were 2 interface areas on the surface of RCγδ in contact with α2A ( Fig 2B–2D ) . First , the larger interaction site ( 715 Å2 ) consisted of 2 adjacent patches of 3 residues each on the RCδ subunit , K59-Y60-K101 ( Fig 2C ) , and R92-Y94-K114 ( Fig 2D ) , which were largely hydrophilic . Second , a smaller hydrophobic site ( 280 Å2 ) on the RCγ subunit consisted of the triad L66-R109-W110 that interacted with helix 3 , helix 4 , and loop 2 of α2A ( Fig 2B ) . Two complementary contact surfaces on the α2A domain extended down from helix C and the metal ion-binding site ( top face ) to the loop 2 sequence S214QYGGD219 ( lateral face ) to form an almost contiguous interface that interacted with the RCγδ subunit . The top face of α2A was approached by the RCδ subunit with its larger 2 patches containing interface ( Fig 2C and 2D ) . The first patch comprised residues K59 , Y60 , and K101 of RCδ interacting with residues D292 and T293 together with the adjacent helix C of α2A . The side chains of K59 and Y60 were countered by complementary carboxylate and hydroxyl groups of D292 and T293 of α2A , while the amino group of K101 pointed towards the backbone carbonyl groups at the C-terminus of helix C . The second patch had the side chains of R92 , Y94 , and K114 of the RCδ subunit pointing into the collagen binding crevice of α2A . The long side chain of K114 of this protuberance sat at the entrance to the divalent cation binding site ( Fig 2D ) and was positioned 7 . 7 Å above the magnesium ion , whereas the positively charged guanidino group and the phenolic hydroxyl group of R92 and Y94 contacted the main chain carbonyl of D219 in loop 2 of α2A . The second contact surface is the loop 2 sequence S214QYGGD219 at the lateral face of α2A , which interacted with the amino acid side chains of L66 , R109 , and W110 of the RCγ subunit ( Fig 2B ) . For example , the aromatic indole ring of W110 contributed to a hydrophobic surface and interacted with the backbone chain of the glycine residues G217 and G218 together with the adjacent aspartate residue D219 within loop 2 of the α2A domain ( Fig 2B ) . In addition , L66 of RCγ contacted N154 of loop 1 of the α2A domain . The final RCγ residue of the triad R109 made contact with the S214 side chain of α2A . Taken together , the hydrophobic patch of the RCγ subunit predominantly interacted with the loop 2 sequence S214QYGGD219 of α2A . This loop 2 sequence immediately preceded residue T221 , which was part of the metal ion binding site of α2A . A key residue with regard to the interface between the RCγδ subunit and the α2A domain in the RCγδ-α2A complex was the loop 2 D219 of α2A , as it was part of both RC contact sites . In addition , it connected the loop 2 sequence with the collagen binding crevice and helix C of α2A . The presence of helix C in the RCγδ-α2A complex structure indicated that RC had trapped the α2A domain in the “closed” conformation , which is not capable of binding collagen [15] . To test whether RC exclusively binds the closed conformation of α2A , we generated 2 conformationally distinct mutants in which the A-domain was held by a disulfide bridge between K168C-E318C and K168C-A325C in the open and closed conformations , respectively ( S2 Fig ) [17 , 41] . Before introducing cysteine residues at these positions , it was necessary to replace the naturally occurring original cysteine residues at position 150 and 270 with alanines . No change in binding affinity to RC was observed for this α2A-C150A , C270A double mutant . In this cysteine-free α2A domain , K168 of α-helix 1 was replaced by a cysteine residue , with a second cysteine residue introduced into α-helix 7 at either position E318 or A325 . As a consequence of the newly formed disulfide bridge , the movement of helices 1 and 7 with respect to each other that occurs when α2A shifts between the “open” and “closed” conformation was blocked . Thus , the α2A domain was held in the “open” ( K168C-E318C ) and “closed” ( K168C-A325C ) conformation , respectively . The α2A mutant with the “open” conformation hardly bound to RC ( Fig 3A ) , while RC binding to the “closed” conformation of α2A ( Kd-value: 0 . 21 ± 0 . 03 nM ) was similar to that obtained with wild-type α2A ( Kd-value: 0 . 29 ± 0 . 02 nM ) . Our structural findings revealed that the sidechain moiety of Lys101 is oriented towards the negatively charged dipole of helix C , stabilizing the closed conformation of the α2A domain ( Fig 3B ) . Among several monoclonal antibodies raised against the RCγδ subunit [40] , IIIG5 belonged to the subgroup that only recognized its epitope within RCγδ after its complexation with α2A and the subsequent release of the RCαβ subunit ( Fig 4A ) . This became evident when the antibody was immobilized and its ability to capture RCαβγδ , RCγδ- α2A , or RCγδ out from solution was probed . IIIG5 gave a binding signal with the RCγδ- α2A complex and RCγδ but not with the RC tetramer alone . Of the 2 RC species capable of binding the IIIG5 antibody , the RCγδ subunit gave the highest binding signal ( Fig 4A ) . The most probable explanation for these results was that the IIIG5 epitope was fully accessible in RCγδ , and so , we observed what approximates the maximal binding . At the other extreme , we had no binding of RCαβγδ , as the epitope was entirely masked in the tetramer . Between these 2 extremes was the RCγδ-α2A complex , in which the epitope is sufficiently exposed for IIIG5 to bind but not to the same extent as for RCγδ due to the nature of the RCγδ-α2A interaction . The sequence epitope of IIIG5 was isolated from a tryptic digestion of RCαβγδ by affinity chromatography on an IIIG5 column and subsequently by reversed-phase high-performance liquid chromatography ( HPLC ) . Mass spectrometry ( MS ) identified the γ chain sequence 94–106 as the IIIG5 epitope ( S3 Fig ) , which was mainly located within the index finger loop of RCγ ( Fig 4B ) . This result can be clearly explained by comparing the native RCαβγδ structure with the newly determined RCγδ-α2A complex structure . The IIIG5 epitope was covered by the RCαβ subunit in the RCαβγδ structure and only became accessible upon formation of the RCγδ-α2A complex . Moreover , the index finger loop of the RCγ underwent a major conformational change upon formation of the RCγδ-α2A complex , leading to increased accessibility of the IIIG5 epitope . The dramatic conformational changes that took place within the RCγδ subunit were readily apparent upon comparing the molecular structures of the RCγδ-α2A complex with the native RCαβγδ tetramer ( Fig 5 ) . The binding face of RCαβγδ changes from a flat surface into a concave binding surface to embrace the α2A domain ( Fig 5A and 5B ) . This was implemented via ( i ) a rigid body movement of both core segments of chains γ and δ , ( ii ) a dramatic re-orientation of the index finger loop of the γ subunit , which harbors the IIIG5 epitope , and , consequently , ( iii ) local re-orientations of key binding residues in both RC subunits ( Fig 5C and 5D ) . The rigid body arrangement can best be described as a flipping of helices 1 and 2 between the RCγ and RCδ subunits whilst maintaining the same relative orientation of the 2 helices within their respective core domains . An additional consequence of this rigid body movement is a conformational shift of the connecting finger loop to track the motion of the opposing core domain . As a result , the 2 core domains flipped over with respect to each other and bent towards the α2A domain to form a concave binding surface such that the RCγδ residues involved in α2A binding were brought into the correct orientation for binding the α2A domain . The apical ends of the index finger loops were in close contact with the CLRP core element of the opposite subunit , forming the 2 interfaces: loop γ–core δ and loop δ–core γ . Whereas the former hardly changed ( Fig 5E and 5F ) , the latter showed a dramatic shift within the RCγδ-α2A complex as compared to the RCαβγδ tetramer ( Fig 5C and 5D ) . In the loop δ–core γ interface of the RCαβγδ tetramer ( Fig 5C ) , a tryptophan core composed of 3 residues ( W76δ , W71δ , and W116γ ) together with a salt bridge between R92δ and D74γ stabilized the index finger loop of the RCδ subunit and oriented it towards the RCγ subunit core sequence connecting helices 1 and 2 . However , in the RCγδ-α2A complex , the salt bridge between R92δ and D74γ found in the RCαβγδ tetramer ( Fig 5C ) was broken . R92δ now formed a hydrogen bond to the main chain of D219 in the α2A loop 2 , and a new salt bridge was observed between R75γ and E77δ and D81δ ( Fig 5D ) . In addition , the RCδ subunit index finger loop became embedded within the antiparallel sheet S3–S4–S5 of the RCγ core such that the indole moiety of W76δ now made van der Waals contacts to Q105γ and Y118γ ( see inset Fig 5C and 5D ) . As a result of these enormous conformational changes , especially at the loop δ–core γ interface , the rigid cores of the 2 RCγδ subunits swung towards each other by about 40°–50° around a hinge located in the center of the index finger swap domain between the cores . This global movement had 2 major consequences . First , as the RCδ subunit snapped into its new position , the 3 key residues of RCγ ( L66 , R109 , and W110 ) underwent a local conformational change that transformed them into an orientation that is competent for α2A binding ( Fig 6A ) . Second , as a consequence of the index finger loop tracking the movement of the RCγ subunit , the contact site between the RCα and RCγ subunits changed its 3D structure due to the formation of the new salt bridge between R75γ and E77δ and D81δ ( Fig 5D ) . Consequently , the previous interface between the RCγ subunit ( K77EQQC81 ) and the RCα subunit ( N74KQQR78 ) became sterically blocked [26] . The movement of the RCγ subunit would also produce steric clashes with the RCβ subunit , and it is likely the combination of these 2 events that resulted in the dissociation of the RCαβ subunit from its RCγδ counterpart . In contrast , the contact site within the RCδ subunit would allow integrin binding irrespective of the conformational change of RC , as their local positions and orientations remained almost unchanged ( Fig 6B ) . In fact , the distance between Y60δ and Y94δ within the RCδ contact sites only changed slightly , from 21 . 7 Å to 20 . 4 Å ( Fig 6B ) , while their distances towards W110γ of the RCγ contact site were reduced from 47 . 5 Å to 31 Å and from 28 . 4 Å to 18 . 6 Å , respectively when comparing the structure of RCαβγδ and RCγδ-α2A complex . This illustrated how significant a reorganization of the RCγδ is required to facilitate the formation of the ultimate inhibitory RCγδ-α2A complex . Unlike helix C , the docking site S214QYGGD219 did not change its conformation between the “open” and “closed” conformation of the α2A domain . To analyze its role , we challenged RC binding to α2A with the monoclonal antibody JA202 . Its epitope had previously been mapped to the sequence QTS214QY [42] and thus overlapped with the RCγ subunit docking site . Among different antibodies against distinct epitopes within α2A , JA202 was the only monoclonal antibody which sterically inhibited RC binding to the α2A domain in a dose-dependent manner ( Fig 7A ) . A comparison of integrin α2 chains from different species showed a high interspecies homology of the loop 2 sequence , S214QYGGD219LT221 ( S4 Fig ) . In contrast , this sequence was absent in A-domains of other integrin α subunits , suggesting that it served as a selective docking site for RC on α2β1 integrin ( S5 Fig ) . Therefore , we replaced the α2A sequence S214QYGGD219L with the corresponding sequence VGRGGRQ of the α1A-domain and tested binding of RC to this α2A-L2α1 mutant . Although this α2A mutant was still able to bind RC , the binding affinity was reduced , as indicated by an increase of the Kd-value from 0 . 76 ± 0 . 12 nM to 2 . 70 ± 0 . 39 nM ( Fig 7B ) . In parallel to the α2A-L2α1 mutant , we exchanged residues in the loop 2 that interacted with RC ( Fig 7C ) , specifically S214 , Y216 , and D219 , as well as the G217 and G218 that are conserved in both integrin α1 and α2 loop 2 sequences , to see which residues were functionally important for the RCγδ-α2A binding . The S214G and D219A mutants , which are located at the outer edges of loop 2 , gave Kd values of 0 . 77 ± 0 . 32 nM and 5 . 2 ± 1 . 36 nM , respectively , while the Y216G mutant in the center of the loop gave a Kd value of 1 . 98 ± 0 . 64 nM ( Fig 7D and 7E ) . In contrast , mutating either of the conserved glycine residues of loop 2 by generating G217K and G218L resulted in a complete loss of RC binding ( Fig 7D ) . This result is in agreement with our structure findings ( Fig 7C ) , which showed that anything larger than a glycine at either position 217 or 218 would sterically clash with the indole side chain of W110γ . In addition , we chemically modified the solvent-exposed W110γ of RC with 2-nitrophenyl sulfenylchloride ( NPS-Cl ) , which introduced a bulky 2-nitro-phenylsulfenyl ( NPS ) group onto the indole side chain . The modified W110γ is no longer able to stack above the 2 glycines G217 and G218 , causing a loss of RC binding to the α2A domain ( Fig 7F ) . Taken together , these results demonstrated that the interaction of W110 of RCγ and the loop 2 of α2A is highly specific and essential for the formation of the high-affinity and inhibitory RCγδ-α2A complex .
Our study reveals not only the interaction sites within RC and its molecular target , the integrin α2A domain , but also the conformational changes that take place within the RCγδ subunit upon α2A binding and the relevance of the 2 contact sites within α2A for RCγδ binding . Moreover , these data suggest a molecular mechanism for the avid and selective interaction of this CLRP and its target . CLRP dimers recognize other target molecules , such as factor IX/X , and the A-domain of vWF by forming a bay region with their joint index finger loop swap domain and 2 flanking core domains . This concave face shapes the binding sites for clotting factors IX and X [43 , 44] and the vWF-factor A-domain [45] . Due to their importance in hemostasis , clotting factors and vWF are valid targets for CLRPs from snake venoms . Bitiscetin and botrocetin interact with the vWF–A1 domain without or together with the glycoprotein Ib ( GPIb ) receptor [27 , 45 , 46] . These studies showed that these snake venom toxins can approach the A-domain from different orientations [35 , 45 , 46] . In yet another orientation , EMS16 approached the α2A domain of α2β1 integrin , which is homologous to the vWF–A1 domain , along its top face directly above the metal binding site and collagen binding crevice , thus preventing collagen from binding [27] . EMS16 and RC are the 2 α2β1 integrin-binding CLRPs whose crystal structures in both the unliganded and the CLRP in complex with the A-domains have been resolved so far [26 , 47] . Although RC approached the α2A domain in a similar orientation to EMS16 , our data revealed that RC , in contrast to any known CLRP structure [27 , 45 , 46] , undergoes a dramatic conformational change to form a concave binding surface . In contrast , the heterodimeric EMS16 did not alter its molecular structure upon α2A binding [27 , 47] , as the concave binding surface required for α2A binding was already preformed . This difference in mode of α2A binding between EMS16 and RC is determined by the distinct quaternary structures of the dimeric EMS16 versus the tetrameric RC and/or by the different purification protocols . When we employed the same purification procedure for RC as for EMS16 and other CLRPs [28–30 , 48] using reversed phase chromatography performed in 0 . 1% trifluoroacetic acid ( TFA ) solution , the RC tetramer dissociated into its subunits α , β , and γδ [49] . The RCγδ subunit alone was still able to bind α2A and to block α2β1 integrin-mediated platelet aggregation specifically [50] , albeit with a different kinetics [40] . Only when applying a milder purification protocol could we obtain a stable RC tetramer and the RCγδ-α2A complex , whose different conformational structures are presented here . Our crystal structure of the RCγδ-α2A complex reveals a geometry of interaction similar to the α2A-bound EMS16 , suggesting that the α2β1 integrin-blocking CLRPs may have a more uniform binding mechanism than the vWF binding CLRPs ( Fig 8 ) . Both CLRPs share the same 2 contact sites within the α2A domain: the conformationally stable loop 2 sequence ( Fig 8C ) and the helix C of the “closed” conformation ( Fig 8D ) . Helix C is recognized by the structurally robust contact area of the RCδ subunit or the homologous EMS16 subunit β ( or B ) . Apart from slight variations of the K59δ side chain and the loop 2 Y216 side chain ( Fig 8D ) adopting an alternate conformation to form a hydrophobic interaction with L66γ , the structures of both complexes are almost identical in this region . In our studies , the role of the loop 2 sequence S214QYGGD219 was reinforced by the JA202 antibody , whose epitope overlaps with this loop 2 sequence and inhibits RC binding completely , presumably due to steric hindrance by the bulky antibody . More subtly , recombinant exchange of the respective loop 2 sequence with the homologous sequence of integrin α1 showed that the loop 2 sequence changes the affinity of the venom component towards the integrin α2 subunit . Similar reductions in the affinity of RC for α2A were also observed with the loop 2 mutants Y216G and D219A . However , a loss of binding was obtained with the G217K and G218L mutants . These 2 glycine residues form part of a shallow dimple on the α2A surface that is covered by W110 of the RCγ subunit . In the molecular structure of the RCγδ-α2A complex , there is not any space to accommodate anything larger than a glycine at either of these 2 positions , which explains the loss of function of these 2 mutants . The loop 2 sequence of the integrin α2A domain is evolutionary conserved between different animal species , especially the GG motif at positions 217 and 218 , but varies remarkably between other integrin α subunits . This suggests that RC’s specificity is mediated by the integrin α2-specific loop 2 sequence , as RC affects α2β1 integrin-mediated platelet blockage in various potential preys but does not affect biological functions mediated by other integrins . Our conclusion—that this cluster of RCγ W110 and G217/G218 of the α2A loop 2 sequence is a key to the RCγδ-α2A interaction—is further supported by the fact that the RC binding is completely lost if the bulky chemical adduct of 2-nitrophenylsulfenyl is introduced to the indole side chain . It is noteworthy that the loop 2 sequence is also relevant for collagen binding , as it forms a hydrophobic contact for the phenylalanine side chain of the middle strand of the trimeric integrin recognition motif of collagen [15] , albeit not as close a contact as with the RCγ W110 side chain . Based on our findings , we suggest the following mode of action ( Fig 9 ) . RCαβγδ interacts with helix C of the α2A domain through the RCδ subunit , where the interacting residues are already in binding-competent orientation . This stabilizes the “closed” conformation of α2A . As a consequence of the movement of RCγ , the RCαβγδ tetramer changes conformation such that RCαβ dissociates from the heterotetrameric assembly . Coupled to this dissociation is the reorganization of L66 , R109 , and W110 of RCγ to interact with loop 2 sequence S214QYGGD219 . Having established both interaction sites , RCγδ firmly binds to α2A and holds it in the “closed” conformation , thereby blocking collagen binding and antagonistically turning off α2β1 integrin signaling . After its release upon formation of the high-affinity RCγδ-α2β1 complex , the RCαβ subunit plays another important role in blocking GPIb and , consequently , vWF-induced platelet aggregation [49] . Moreover , our biochemical data showed that the RCαβ subunit is significantly more soluble than the RCγδ subunit [40] . Therefore , it likely acts as a solubility enhancer to ensure that the RCγδ subunit is delivered to α2β1 integrin . Once RCγδ has bound to its target and the RCαβ subunit has been released , RC effectively shuts down the 2 platelet receptors , α2β1 integrin and GPIb , thereby effectively blocking both collagen-induced and vWF-induced platelet activation and aggregation . In summary , a comparison of the RCγδ-α2A structure with the EMS16-α2A integrin complex [27] shows that the residues involved in the binding of RC and EMS16 to α2β1 integrin are highly conserved . The formation of the inhibitory RC-α2A complex requires both the interaction of RCδ with the helix C of α2A and RCγ with the α2A loop 2 sequence . Furthermore , the presence of helix C in our structure confirms that we have trapped α2A in the “closed” conformation , which is not able to bind collagen and explains why RC is able to block collagen-mediated platelet aggregation . Finally , the requirement of 2 separate sites within the α2A domain for both function and specificity may be instrumental for the design of novel α2β1 integrin inhibitors .
RC and its γδ subunit were isolated as previously described [40 , 51] . The monoclonal antibodies ( mAbs ) against RC , among them IIIG5 from mice and IC3 from rats , were generated and isolated as previously described [40] . The murine mAbs against the human α2A domain , JA202 and JA218 , were a generous gift from D . Tuckwell ( formerly of the University of Manchester , United Kingdom ) [40 , 42] . PCR primers were obtained from Eurofins ( Eurofins Genomics , Germany ) and are written in 5′-3′ direction . Restriction enzymes and molecular biology reagents were from Thermo Fisher Scientific ( Germany ) unless otherwise stated . Cloning products and expression vectors were validated by DNA sequencing ( Eurofins Genomics ) . RC , dissolved at 110 μM in 30% acetic acid solution , was treated with 9 . 2 mM 2-nitrophenyl sulfenylchloride ( NPS-Cl , TCI Chemicals , Germany ) or left untreated for 1 h at 20 °C in the dark according to [52] , subsequently dialyzed against 0 . 1% TFA ( RP-solution ) and separated on a Supercosil C18 column ( Supelco , Germany ) by reversed-phase chromatography as described [26] . The RCγδ-containing fractions were pooled , lyophilized , and dissolved in RP-solution containing 30% acetonitrile . Purity was assessed by SDS-PAGE . Spectroscopic evaluation at 365 nm according to [52] confirmed the covalent modification of RC tryptophan residues with 2-nitro-phenylsulfenyl ( NPS ) -groups . The His6-tagged α2A domain was generated as previously described [26 , 53] . It was loaded onto a HiTrap Ni Sepharose column ( GE Healthcare; 5 ml ) previously equilibrated with PBS/MgCl2-buffer , pH 7 . 4 ( 20 mM sodium phosphate , pH 7 . 4 , 150 mM NaCl , 1 mM MgCl2 ) . After washing with the same buffer , the RCαβγδ-containing fractions from the RC isolation with MonoS column [51] were applied to the α2A domain loaded Ni Sepharose column after having been treated with 0 . 5 μM phenylmethylsulfonyl fluoride ( PMSF ) and 1 μg/ml aprotinin to prevent proteolytic digestion by potentially contaminating snake proteases . After RCαβγδ had bound to the Ni Sepharose-immobilized α2A domain , the HiTrap Ni Sepharose column was washed with PBS/MgCl2-buffer , pH 7 . 4 . Then , the column was washed with PBS/EGTA-buffer , pH 7 . 4 ( 5 mM EGTA in 20 mM sodium phosphate , pH 7 . 4 , 150 mM NaCl ) and the RCαβ subunit eluted . After another washing step with PBS/MgCl2-buffer , pH 7 . 4 , the RCγδ-α2A complex was eluted with a linear gradient of 0–200 mM imidazole in PBS/MgCl2-buffer , pH 7 . 4 from the HiTrap Ni Sepharose column . Protein concentration in the imidazole eluate was determined using the Bradford reagent ( BioRad ) . For crystallization , the complex-containing fractions were pooled and digested with TPCK-treated trypsin ( Sigma-Aldrich ) at an enzyme:substrate ratio of 1:100 at 37 °C for 1 h . The digest was stopped with 1 mM PMSF , concentrated and separated by gel filtration to remove excess α2A domain , trypsin and contaminating peptides from the RCγδ-α2A complex . The TSK G2000SWXL chromatography was performed in 10 mM HEPES , pH 7 . 4 , 100 mM NaCl buffer . The RCγδ-α2A complex was concentrated by ultrafiltration and its protein concentration determined with the Bicinchoninic Acid Protein Assay ( BCA , Thermo Fisher Scientific ) . To analytically prove the physical contact of both partners , the complex was cross-linked with 0 . 5 mM bi-sulfosuccinimidyl-suberate ( BS3 , Thermo Fisher Scientific ) . Its IEP was determined to be pH 6 . 5–6 . 8 and pH 6 . 7 by isoelectric focusing in precast ZOOM pH 3–10 gels ( Thermo Fisher Scientific ) and by analytical chromatofocusing on a MonoP column ( GE HealthCare ) with a pH gradient of 7 . 4 to 4 . 0 , respectively . Crystals of 10 mg of RCγδ-α2A were grown by hanging-drop vapor diffusion at 293 K by mixing 2 μL of protein solution with 2 μL reservoir solution containing 2 . 65 M ammonium sulfate and 100 mM Tris pH 8 . 0 . Crystals appeared after 6 weeks and were soaked in mother liquor containing 20% glycerol for 5–10 min before being flash frozen in liquid nitrogen . Diffraction data was collected at the Canadian Light Source CMCF-08ID-1 beamline ( λ = 0 . 97949Å ) at 100 K using a Rayonix MX225 CCD detector . The dataset was indexed , integrated , and scaled with MOSFLM [54] and the CCP4-package [55] . The spacegroup is P41 with 6 molecules in the asymmetric unit ( see also Table 1 ) . The phases were determined by rigid body refinement using the previously solved RC structure ( PDB code 3GPR ) in Refmac [56 , 57] . The model was built and refined without NCS restraints using Coot [58] and refined with the Phenix software package [59] . The crystallographic data and refinement statistics are summarized in Table 1 . The final coordinates and structure factor amplitudes were deposited in the PDB ( RCSB-code: 5THP ) . The human α2A domain and its mutants were produced in a bacterial expression system . The expression vectors encoding the disulfide-locked conformation mutants of α2A were generated using a previously described pET15b-His6-α2A construct ( residues 142 through 337 of human integrin α2 ) . To replace the endogenous cysteine residues at 150 and 270 , this plasmid was used as template for a 2-step PCR with the 3 primer pair sets ( i ) HTfwd ( CTCTCCATGGGCTCTTCTCATCATCATCATCATCATTC ) and R1 ( C11A ) ( CATCAGCCACAACCACAAC ) , ( ii ) F2 ( C11A ) ( TTGTGGCTGATGAATCAAATAG ) and R2 ( C131A ) ( TTGGCTTGATCAATCACAGC ) , and ( iii ) F3 ( C131A ) ( ATTGATCAAGCCAACCATGAC ) and α2Arev ( CGGACATATGCTAACCTTCAATGCTGAAAAATTTG ) in the first set of reactions . The 3 amplicons were purified and again PCR-amplified with the outer primer pair HTfwd and α2Arev to a 670 bp amplicon , which , after A-tailing with Taq DNA polymerase , was intermediately ligated into pCR2 . 1 TOPO , excised with NdeI and NcoI , and the restriction fragment was subcloned into the linearized , NdeI , NcoI-cleaved pET-15b expression vector . The final expression plasmid pET-15b-His6-α2A ( C150 , 270A ) was transformed into Escherichia coli BL21 ( DE3 ) . To generate the disulfide-locked conformation mutants of α2A , which share the same K168C mutation but differ in E318C ( “open” conformation: K168C , E318C ) or A325C ( “closed” conformation: K168C , A325C ) , 3 rounds of PCR amplification were performed . In the first , site-directed mutagenesis K168C was introduced by amplifying the entire plasmid with the back-to-back primer pair K168C fw ( AAGGCCTGGATATAGGCCCC ) and K168C rev ( GTACAAAGCATTCCAAAAAATTCTTTACTGC ) . Based on this mutation , the final 2 mutants ( K168C , E318C; K168C , A325C ) were similarly generated using the primer pairs E318C fw ( GTCTGATTGCGCAGCTCTACTAGAAAAG ) /E318C rev ( ACATTGAAAAAGTATCTTTCTGTTGGAATAC ) and A325C fw ( ATTAGGAGAACAAATTTTCAGCATTGAAG ) /A325C rev ( GTCCCGCACTTTTCTAGTAGAGCTG ) . For each site-directed mutagenesis , only 1 primer contained the specific mutation . The PCR products were amplified by the Phusion Hot Start II polymerase and covered the whole template vector ( 6307 bp ) with the mutation . After the original , methylated vector had been digested with DpnI , the amplicons were purified using the DNA Clean & Concentrator Kit ( Zymo Research ) , followed by 5′-phosphorylation with T4 polynucleotide kinase and religated using T4 DNA ligase . For protein expression , E . coli strain BL21 ( DE3 ) were transformed with the validated plasmid constructs encoding the α2A domain in its “open” ( pET-15b-His6-α2A-C150/270A-K168C/E318C ) and “closed” ( pET-15b-His6-α2A-C150/270A-K168C/A325C ) conformations . The α2A-L2α1 mutant , in which the sequence S214QYGGDL is replaced by the corresponding loop 2 sequence V214QRGGRDQ of the integrin α1 A-domain , was generated by 2-step PCR . The pET15b-construct encoding the His-tagged α2A domain [26] was used as a template . The primer pairs α2A fw ( GGATATCTGCAGAATTCGCCCTTC ) and R1_a1insert into a2 ( CTTTACTAACATCGTTGTAGGGTCTGTCACGTCGCGCCACCAGCGGTC ) , F1_a1insert into a2 ( GTGCAGCGCGGTGGTCGCCAGACAAACACATTCGGAGCAATTC ) , and α2A rev ( AGGCCATATGCTAACCTTCAATGCTGAAAATTTG ) amplified the N- and C-terminal halves of the cDNA . The 2 amplicons were mixed and amplified with the outer primer pair . The resulting 680 bp amplicon was trimmed with NcoI and NdeI , ligated into a correspondingly cut pET-15b vector , verified by sequencing , and transformed into E . coli BL21 ( DE3 ) . Point mutations within the loop 2 sequence were also generated by a 2-step PCR using the wild-type α2A-encoding cDNA as template . First , cDNA fragments encoding the N- and C-terminal halves of α2A were amplified by using the 2 pairs of forward outer and reverse inner primers and of forward inner and reverse outer primers , respectively , as summarized in Table 2 . The amplicons were purified and taken as template for a second PCR with the outer primer pair to obtain the wild-type and mutant α2A domains encoding cDNAs , which were digested with NdeI and BamHI and ligated into the likewise-cut pET-15b vector . After verification by sequencing , the expression vectors were transformed into E . coli BL21 ( DE3 ) . All α2A domain mutants were purified using HiTrap Ni Sepharose column ( GE HealthCare ) as per the wild type . The wells of a half-area microtiter plate ( Costar ) were coated with 10 μg/ml His-tagged α2A domain in TBS/Mg buffer ( 50 mM Tris/HCl , pH 7 . 4 , 150 mM NaCl , 3 mM MgCl2 ) at 4 °C overnight . After washing twice with TBS/Mg buffer , the wells were blocked with 1% BSA in TBS , pH 7 . 4 , 2 mM MgCl2 for 1 h at room temperature . The immobilized α2A domain was titrated with a serial dilution of RCαβγδ or RCγδ without and with NPS-modified tryptophans in the blocking buffer for 1 . 5 h . For the mAb inhibition experiment , RC at a constant concentration of 2 nM was added to the wells in either the absence or presence of mAb JA202 against RC . After washing twice with HEPES-buffered saline ( HBS ) ( 50 mM HEPES/NaOH , pH7 . 4 , 150 mM NaCl , 2 mM MgCl2 ) , bound RC was fixed with 2 . 5% glutaraldehyde in the same solution for 10 min at room temperature . After 3 additional washes with TBS/Mg buffer , bound RC was quantified by ELISA using a primary rabbit antiserum against RC and a secondary alkaline phosphatase conjugated anti-rabbit–IgG antibody , each diluted 1:2 , 000 in 1% BSA/TBS/Mg . Conversion of para-nitrophenyl phosphate ( pNpp ) to para-nitrophenolate was stopped with 1 . 5 M NaOH and measured at 405 nm . The titration curves were evaluated as described below . The inhibition curves were approximated by GraphPad Prism software using the inhibition vs . log [inhibitor]-approximation . To compare independent inhibition and binding experiments , the dynamic ranges were normalized to the mAb-free control and to the saturation value of the wild-type α2A domain , respectively . Alternatively , the α2A domains , either wild-type or mutants , were captured using the mAb JA218 at a ligand-binding–irrelevant epitope , thereby avoiding any conformational changes due to adsorption to the plastic . To this end , 2 . 5 μg/ml JA218 was immobilized to a microtiter well at 4 °C overnight . After the wells were washed twice with TBS/Mg buffer , wells were blocked with 1% BSA in the same buffer for 1 h , and then , the α2A domain was added at 10 μg/ml for 1 h . After washing the wells , RC was titrated and detected as described above . The mAb IIIG5 was coated to the wells of a microtiter plate at 3 μg/ml in TBS/Mg buffer overnight . After 2 washing steps , wells were blocked with 1% BSA in TBS/Mg buffer for 1 h and then titrated with either RCαβγδ , RCγδ , or RCγδ-α2A complex for 1 . 5 h at room temperature . Bound RC was fixed and quantified as described above . A mathematical approximation of the titration curve , including determination of Kd-values , is described below . IIIG5 was immobilized to cyanogen bromide-activated sepharose according to the manufacturer’s instruction ( GE Healthcare ) . RCαβγδ-containing fractions from the Mono S purification of C . rhodostoma venom [26] were reduced with 4 mM tris ( hydroxymethyl ) phosphine ( THP , Calbiochem ) for 20 min at 60 °C , and free thiol groups were alkylated with 16 mM iodoacetic acid . The protein was precipitated with trichloroacetic acid , washed with acetone twice , resuspended in 87 . 5 mM sodium bicarbonate/0 . 5 M urea and digested with TPCK-trypsin for 23 h at 37 °C . After addition of 1 mM PMSF , the digest was diluted with TBS/HCl buffer , pH 7 . 4 and loaded onto the IIIG5 column . The RC peptide harboring the IIIG5 epitope was eluted in a pH gradient from pH 7 . 5 to 3 . 0 in 20 mM citrate buffer and further purified by reversed phase on a Supercosil C18 column in a 0%–28% acetonitrile gradient in 0 . 1% TFA/water . Lyophilized HPLC fractions were dissolved in 40% methanol containing 0 . 5% formic acid and analyzed by nano-electrospray ionization ( nanoESI ) MS and MS/MS . Peptide structures were deduced from the corresponding fragment ion spectra . NanoESI MS experiments were carried out by using a SYNAPT G2-S mass spectrometer ( Waters , Manchester , UK ) equipped with a Z-spray source in the positive ion sensitivity mode . Typical source parameters were as follows: source temperature , 80 °C; capillary voltage , 0 . 8 kV; sampling cone voltage , 20 V; and source offset voltage , 50 V . For low-energy collision-induced dissociation ( CID ) experiments , the peptide precursor ions were selected in the quadrupole analyzer , subjected to ion mobility separation ( IMS; wave velocity 850 m/s , wave height 40 V , nitrogen gas flow rate 90 ml/min , and helium gas flow rate 180 ml/min ) , and fragmented in the transfer cell using a collision gas ( Ar ) flow rate of 2 . 0 ml/min and collision energies up to 100 eV ( Elab ) . In titration curves , a signal S , usually the extinction at 405 nm caused by the alkaline phosphatase-catalyzed conversion of pNpp , is measured in response to the total concentration c0 of added titrant . Based on a Michaelis–Menten-like binding mechanism , we deduced the following equation to approximate titration curves , if the signal S and the total concentration c0 of added ligand ( RC ) is known: S ( c0 ) = ( SM−Sm ) ∙ ( ( c0+cR+K ) − ( c0+cR+K ) 2−4∙c0∙cR2∙cR ) +Sm+B∙c0 with SM and Sm , maximum and minimum signals , respectively; cR , the concentration of ligand binding site ( equals the receptor concentration for monovalent receptors ) ; and K , the dissociations constant Kd . The term B·c0 takes into account a linear change in the signal due to nonspecific binding of the ligand . The 5 parameters SM , Sm , cR , K , and B are calculated by nonlinear regression from titration curves . The data from titration and inhibition curves were statistically evaluated using GraphPad Prism software . Values were usually compared with the values obtained for the wild-type α2A or nonmodified RC with Student t test , where the significance level was set at 1% unless otherwise stated . | In animals , collagen-mediated platelet aggregation is an essential component of the blood’s clotting response following vascular injury . A small group of snake venom toxins belonging to the C-type lectin protein family exert their harmful effects by directly targeting this pathway . Rhodocetin ( RC ) is a heterotetrameric protein found in the venom of the Malayan pit viper ( C . rhodostoma ) . RC specifically binds α2β1 integrin , the key protein required for collagen-mediated platelet aggregation . In this study , we describe the interaction between RC and α2β1 integrin at atomic resolution . This study reveals that RC undergoes a massive structural reorganization upon α2β1 integrin binding , such that RC’s αβ subunit is released from its γδ subunit and a γδ-α2β1 integrin complex is formed . The inhibitory nature of this complex can be readily explained as RC binding along the top surface of the α2β1 integrin and directly above the collagen binding site . As a result , access of collagen to its binding site is blocked , thereby preventing collagen-mediated platelet aggregation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"toxins",
"crystal",
"structure",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"condensed",
"matter",
"physics",
"organic",
"compounds",
"toxicology",
"toxic",
"agents",
"collagens",
"mutation",
"integrins",
"amino",
"acids",
"sequence",
"motif",
"analysis",
"crystallography",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"venoms",
"sequence",
"analysis",
"glycine",
"solid",
"state",
"physics",
"bioinformatics",
"cell",
"adhesion",
"proteins",
"extracellular",
"matrix",
"chemistry",
"physics",
"biochemistry",
"point",
"mutation",
"molecular",
"structure",
"cell",
"biology",
"organic",
"chemistry",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"chemical",
"physics"
] | 2017 | Dramatic and concerted conformational changes enable rhodocetin to block α2β1 integrin selectively |
Chromosomal insertions are genomic rearrangements with a chromosome segment inserted into a non-homologous chromosome or a non-adjacent locus on the same chromosome or the other homologue , constituting ~2% of nonrecurrent copy-number gains . Little is known about the molecular mechanisms of their formation . We identified 16 individuals with complex insertions among 56 , 000 individuals tested at Baylor Genetics using clinical array comparative genomic hybridization ( aCGH ) and fluorescence in situ hybridization ( FISH ) . Custom high-density aCGH was performed on 10 individuals with available DNA , and breakpoint junctions were fine-mapped at nucleotide resolution by long-range PCR and DNA sequencing in 6 individuals to glean insights into potential mechanisms of formation . We observed microhomologies and templated insertions at the breakpoint junctions , resembling the breakpoint junction signatures found in complex genomic rearrangements generated by replication-based mechanism ( s ) with iterative template switches . In addition , we analyzed 5 families with apparently balanced insertion in one parent detected by FISH analysis and found that 3 parents had additional small copy-number variants ( CNVs ) at one or both sides of the inserting fragments as well as at the inserted sites . We propose that replicative repair can result in interchromosomal complex insertions generated through chromothripsis-like chromoanasynthesis involving two or three chromosomes , and cause a significant fraction of apparently balanced insertions harboring small flanking CNVs .
Chromosomal insertion occurs when a segment of one chromosome is translocated and inserted into an interstitial region of another non-homologous chromosome ( interchromosomal insertion ) , or into a different region of the same chromosome ( intrachromosomal insertion ) . Insertions are considered as complex chromosomal rearrangements ( CCRs ) since they require at least three chromosome breakage events . [1] Chromosomal insertions are also considered as complex genomic rearrangements ( CGRs ) as they consist of more than one simple rearrangement , and have two or more DNA breakpoint junctions . [2 , 3] By cytogenetic techniques , the incidence of microscopically visible insertions was estimated to be 1 in 80 , 000 live births . [4] More recently , by array-comparative genomic hybridization ( aCGH ) in conjunction with fluorescence in situ hybridization ( FISH ) confirmation of the aCGH findings , insertion events were demonstrated to occur much more frequently , with estimated incidence of 1 in 500[1] or 1 in 563[5] individuals tested . Another study demonstrated that ~2 . 1% of apparently de novo , interstitial CNVs were actually consequences of imbalances resulted from parents with balanced insertions . [6] These data highlight the importance of identifying such parental genomic information for reproductive counseling and potential recurrence risk estimates . Phenotypic consequences of insertions vary , depending on the size , gene content and orientation of the inserted fragment , in addition to possible disruption or dysregulation of a gene or topologically associating domain ( TAD ) at the inserted genomic locus . Complex insertions are defined as insertions generated by more than three DNA breakages and joining events . [1 , 7] Usually , additional copy-number gain or loss is observed at the inserted site for these events complicating the interpretation of potential phenotypic consequences observed . Little is known regarding the molecular mechanisms for the formation of insertions; particularly with regards to the mechanism ( s ) of formation of complex insertions . Thus , we aimed to elucidate the potential underlying mechanisms generating complex insertions . Surprisingly , we observed complex insertions as part of apparent chromothripsis-like , chromoanasynthesis events involving two or three chromosomes . Chromothripsis was first described as a catastrophic phenomenon in cancer genomes and observed as highly complex somatic rearrangements , with a distinct pattern of frequent oscillations between neutral and deleted copy-number states and seemingly focused on one chromosome . [8] A similar apparent chromosome shattering mechanism has been observed as de novo mutations in individuals with neurodevelopmental abnormalities , and this type of germline chromothripsis involves complex balanced rearrangement among several chromosomes . [9] These events may appear as balanced rearrangements by conventional metaphase chromosome analysis . Both somatic and germline chromothripsis were proposed to be caused by a similar chromosome shattering mechanism that undergoes repair through non-homologous end-joining ( NHEJ ) . [10 , 11] A third type of chromothripsis-like event , defined as chromoanasynthesis , was observed as de novo constitutional CGRs involving region-focused copy-number changes including duplications and triplications . These chromoanasynthesis events were proposed to be generated through replication-based mechanisms , such as fork stalling and template switching and/or microhomology-mediated break-induced replication ( FoSTeS/MMBIR ) with iterative template switching resulting in extensive complexity . [12] The molecular analysis and findings in complex insertions we report here mostly resembled the patterns observed in constitutional genomic chromoanasynthesis events . In this study , we identified 16 individuals with distinct complex insertions among 56 , 000 individuals tested at Baylor Genetics ( BG ) using clinical aCGH and FISH . We fine-mapped DNA breakpoint junctions in 6 complex insertions at nucleotide resolution , and three of them resembled chromoanasynthesis events with multiple chromosomes involved . In addition , we analyzed 5 families with unbalanced insertions detected in probands and inherited from parents with apparently balanced insertion detected by FISH analysis . We found that 3 parents had additional small CNVs at one or both sides of the inserting fragments as well as at the inserted sites likely generated during formation of the structural variant . We propose that these events are due to DNA replicative repair errors generated by replication-based mechanism ( s ) using iterative template switching . [3]
Previously , we demonstrated that by performing confirmatory FISH of the copy-number gains identified in clinical chromosome microarray analysis ( CMA ) testing , some duplications were shown not to be represented by tandem duplication events , but were rather translocated and inserted at another locus in the genome . [1] This approach allowed the discovery of 40 individuals with insertions among the 18 , 000 individuals tested in the CMA laboratory at BG from July 2005 to January 2009 . Among these 40 individuals , 8 were found to carry complex insertions ( S1 Table , individuals Cplex1–8 ) . [1] In this study , we expanded the cohort to 56 , 000 individuals tested from July 2005 to November 2014 , and identified an additional 76 individuals with chromosome insertions ( out of the subsequent 38 , 000 individuals tested ) , therefore resulting in the incidence of insertions being consistently about 1 in 500 . This incidence is likely underestimated given that some of the insertions are too small to be verified by FISH . Among these latter 76 individuals , we identified another 8 individuals with complex insertions ( S1 Table , individuals Cplex9–16 ) . Among the 16 individuals with complex insertions , 2 are intrachromosomal insertions ( Cplex1 and Cplex2 ) , and the remaining 14 are interchromosomal insertions ( Cplex3–16 ) ( S1 Table ) . Cplex2 was previously demonstrated in detail with all proposed breakpoint junctions mapped ( BAB3105 from Ref . 12 ) in the paper that first defined the chromoanasynthesis phenomenon and thus was excluded from the current study . For the remaining 15 individuals , we were able to obtain genomic DNA from 10 individuals ( Cplex1 , 3 , 4 , 5 , 6 , 7 , 9 , 11 , 12 , and 16 ) and repeated the CMA testing using the latest version ( Baylor CMA version 10 . 2 oligo ) . [13] To map the breakpoint junctions to nucleotide resolution , we further designed high-density custom aCGH specifically targeting the inserted fragment and the potential inserting loci in 8 individuals based on the CMA results ( Cplex3 , 4 , 5 , 6 , 7 , 9 , 11 and 12 ) . By long-range PCR with Sanger sequencing , we were able to map ( or partially map ) the breakpoint junctions to nucleotide resolution in 6 individuals ( Cplex4 , 5 , 6 , 9 , 11 and 12 ) . For the remaining 4 individuals , we were unable to map breakpoint junctions , probably due to the limitations of the techniques applied in this study and potential further complexity at those breakpoints . Parental samples were not available for these 16 individuals . Among the 6 individuals with breakpoint junctions mapped , three individuals showed basic complex insertions , with a duplicated fragment translocated and inserted into another genomic locus with a deletion at the inserting position ( Fig 1 , S1 Table ) . Cplex4 demonstrated an ~11 . 8 Mb duplication on chromosome 14 ( 14q22 . 3q24 . 1 ) and an ~4 . 4 Mb deletion on chromosome 13 ( 13q21 . 31q21 . 32 ) revealed by array results ( Fig 1A ) . FISH analysis and breakpoint junctions mapping demonstrated that the third copy of the duplicated segment on chr14 ( chromosome 14 ) was inserted into chr13 at the position of the deletion ( Fig 1B , FISH images previously published in a case report ) . [14] Similarly , in individual Cplex9 , array results revealed an ~2 . 2 Mb duplication on chr9 at band 9q21 . 31 and an ~8 . 3 Mb deletion on chr13 at bands 13q12 . 3q13 . 3 , while in Cplex12 , array results revealed an ~0 . 8 Mb duplication on chr6 at band 6q27 and an ~0 . 5 Mb deletion on chr5 at band 5p14 . 3 . Both FISH and breakpoint junctions mapping demonstrated that the duplicated fragment was inserted into the locus at which the deletion was observed in these latter two individuals Cplex9 and Cplex12 ( Fig 1C and 1D and S1 , S2A and S3A Figs ) . Note that in individual Cplex4 , the inserted fragment was in the same orientation as the reference genome , however , in both Cplex9 and Cplex12 , the inserted fragments were inverted when insertionally translocated ( Fig 1B–1D ) . The CGRs in all three individuals were proposed to be generated through two breakpoint junctions , with 1 bp microhomology observed at both junctions in Cplex4 ( Fig 1A , Table 1 ) , 2 bp and 3 bp microhomologies observed at the junctions in Cplex9 ( S2B Fig , Table 1 ) , and 2 bp and 3 bp small insertions at junctions in Cplex12 ( S3B Fig , Table 1 ) . In contrast to the three individuals described above , Cplex5 , Cplex6 , and Cplex11 showed multiple CNVs in addition to the insertions and were generated through multiple breakpoint junctions ( Table 1 ) . Cplex5 exhibited 4 CNVs from the array results on both chr6 and chrX: an ~1 . 3 Mb duplication at 6q21 , an ~0 . 4 Mb deletion at 6q24 . 2 , an ~8 . 6 Mb deletion at 6q25 . 1q25 . 3 ( resulting in an overall duplication-normal-deletion-normal-deletion CGR pattern on chr6 ) , and an ~1 . 5 Mb duplication at Xq28 ( Fig 2A ) . FISH analysis revealed that the duplicated fragment of Xq28 was inserted and translocated to chr6 , potentially at the deleted locus of 6q24 . 2 ( S1 Table ) . Breakpoint junction mapping confirmed the findings observed by FISH , and the 4 mapped junctions enabled developing a parsimonious model accounting for all available data potentially explaining the rearrangement in this individual ( Fig 2B ) . In brief , the duplicated fragment of Xq28 was inserted into 6q24 . 2 , replacing the deleted region of 6p24 . 2 ( Junction 1 and 2 ) , while a duplicated fragment of 6p21 was inserted into 6q25 . 1 , again replacing the other deleted region of 6q25 . 1q25 . 3 ( Junction 3 and 4 ) . Breakpoint junction sequencing revealed a 7 bp templated insertion ( copied from nearby sequences ) at Junction 1 , 2 bp microhomology at Junctions 2 and 4 , and 5 bp microhomology at Junction 3 ( Table 1; S4 Fig ) . In individual Cplex6 , CMA showed an ~0 . 58 Mb duplication at 5p15 . 33 , and an ~0 . 07 Mb duplication at Xq28 . High-density aCGH revealed that the duplication on Xq28 actually contained a small triplication ( ~6 kb ) embedded in the duplication ( S5A Fig ) . FISH analysis and breakpoint junction mapping demonstrated that the duplicated fragment of 5p15 . 33 was inserted in an inverted orientation to Xq28 . In addition , the triplication was also embedded in the duplication in an inverted orientation ( Fig 2C , S1 Table ) , revealing a duplication—inverted triplication—duplication; a CGR pattern analogous to that previously observed and designated DUP-TRP/INV-DUP . [15] The proximal side of the duplication at Xq28 was joined to the distal side of the duplication at 5p15 . 33 ( Fig 2C , Junction 1 ) , while the proximal side of the 5p15 . 33 duplication was joined to the proximal side of the triplication embedded in the Xq28 duplication , leading to the triplication being inverted ( Junction 2 ) . We hypothesized a third junction connecting both distal sides of the triplication and the duplication at Xq28 should be present to generate the overall CGR in this individual; however , we were unable to uniquely position and map this breakpoint , possibly due to the presence of a low copy repeat ( LCR ) ( S5A Fig ) . Sequences of Junction 1 in this individual showed blunt ends , while Junction 2 showed an insertion of 376 bp templated from at least three nearby genomic loci on both chr5 and chrX ( Table 1 , S5B Fig ) . Individual Cplex11 exhibited the most complicated rearrangement in this study . Array results demonstrated a duplication-normal-duplication-normal-deletion pattern at 13q33 . 2 to 13q34 and a duplication-normal-duplication-triplication-duplication pattern at Xq21 . 1 ( Fig 3A ) ; FISH analysis showed that both of the two duplicated regions on chr13 were inserted into chrX ( S6 Fig , S1 Table ) . Breakpoint mapping further demonstrated that the rearrangement between chr13 and chrX could be potentially generated through 6 junctions ( Fig 3B ) . With the exception of the hypothetical Junction 5 , we were able to map the remaining 5 junctions to nucleotide resolution . Based on the information of the five junctions mapped and the CNVs observed , we proposed the existence of Junction 5 to most parsimoniously explain the observed overall rearrangement in this individual ( Fig 3B ) . Upon careful examination of the junctions , we observed that sequences of Junction 2 contained an 8 , 192 bp insertion from Xq13 . 2 , followed by a 5 , 167 bp insertion from 4q13 . 1 , leading to the discovery of the involvement of a third chromosome , chromosome 4 , in this individual’s CGR ( Table 1 , S7 Fig ) . The remaining mapped junctions showed 2 bp microhomology ( Junction 6 ) or blunt ends ( Junction 1 , 3 and 4 ) . Previously , we reported a child ( BAB1379 ) with PLP1 deletion that resulted from a maternal balanced insertion ( BAB1381 ) of a segment on chrX containing the entire PLP1 gene translocated and inserted into the telomeric region of the q arm of chr19 ( Fig 4A ) . [16] The PLP1 deletion breakpoint junction was mapped in the previous publication , showing an Alu-Alu mediated rearrangement ( Junction 3 in S8 Fig , re-drawn in hg19 ) . This junction was present in the mother ( BAB1381 ) and her affected son with Pelizaeus-Merzbacher disease ( BAB1379 ) , but not in the unaffected son ( BAB1380 ) . To fine map other breakpoint junctions involving the insertion , we designed high-density aCGH targeting both the regions on chrX containing PLP1 , and the potential insertion site at 19qter . Surprisingly , in the mother , we did not see complete copy-number neutral genomic intervals around the PLP1 region as expected for her balanced insertion , but instead observed small CNVs that map at the exact loci of both ends of the deletion position in her affected son ( Fig 4B ) . More specifically , an ~10 kb deletion at the proximal boundary , and an ~22 kb duplication at the distal boundary of the deletion position in her son ( Fig 4B ) . In addition , an ~182 kb duplication was detected at 19q13 . 4 , the potential inserting site , in the mother ( Fig 4B ) . Further breakpoint junction mapping in the mother revealed that the distal side of the duplication on chr19 joined the distal side of the small deletion on chrX ( Junction 1 in Fig 4C ) , while the proximal side of the chr19 duplication joined the distal side of the small duplication on chrX ( Junction 2 in Fig 4C ) . The two small CNVs detected on chrX in the mother were actually due to unbalanced insertion from chrX to chr19 , together with a duplication at the inserting site at 19q13 . 4 . Sequences of the junctions showed 3 bp microhomology ( Junction 1 ) and 15 bp templated insertion from nearby sequences at Junction 2 ( Table 1 , S8 Fig ) . Observations in this family intrigued us to consider that the phenomenon may not be unique—CNVs inherited from parents with apparently balanced insertions may not be completely balanced at the molecular level . Given the small size of the potential CNVs , some may evade detection by clinical CMA . Therefore , we searched for similar families in the CMA database at BG , and found 12 families with a proband having a CNV inherited from a parent with apparently balanced insertion ( named Family 1 to Family 12 ) . We consented 4 families ( Family 3 , 4 , 7 and 12 ) for further research studies , and discovered that in 2 families ( Family 3 and Family 12 ) , the apparently balanced insertions in the parents were not completely balanced , but actually had additional complexities revealed by molecular analyses . In Family 3 , Proband 3 ( P3 ) showed a ~4 . 588 Mb deletion at 7p15 . 2p14 . 3 from array results; this deletion was further found by FISH analysis to be inherited from Mother 3 ( Mat3 ) with apparently balanced insertion from chr7 into 9p24 ( S9A Fig , S1 Table ) . Upon careful interpretation of high-density aCGH results , a small deletion ( ~4 kb ) was observed in the mother at the exact boundary of the deletion in her child ( Fig 5 ) . We were able to fine map the identical deletion breakpoint junction present in both P3 and Mat3 . Interestingly , an 815 bp insertion from 9p24 ( chr9:5874574–5875388 ) was found at the chr7 junction sequences ( Jct1 ) –the potential insertion locus observed from FISH in Mat3 ( Fig 5 ) . We further performed high-density aCGH in both Mat3 and P3 targeting the entire short arm of chr9 . No promising CNVs were identified in either Mat3 or P3 , however , three probes covering chr9:5874574–5875388 showed elevated ratio only in P3 but not Mat3 ( S9B Fig ) . Based on this observation , we suspected an exchange of genetic material between chr7 and chr9 in the mother Mat3 –the ~4 . 588 Mb fragment from 7p15 . 2p14 . 3 was inserted to chr9 , replaced by a small fragment from chr9p24 . 1 ( 815 bp from chr9:5874574–5875388 ) . Note that the large fragment of 7p15 . 2p14 . 3 broke and re-joined during the inserting process based on the observation of mapped breakpoint junction 2 ( Jct2 ) , and additional junctions ( s ) should be present that connect the inserted fragments from chr7 to chr9 except for the mapped junction 3 ( Jct3 , S9C Fig ) . Her child P3 inherited the deleted chr7 with the 815 bp insertion from chr9 , together with an unaltered paternal chr9 . We also suspected that the insertion site on chr9 was around chr9:5874574–5875388 , and therefore performed walk-in PCRs and successfully pinpointed the insertion site on chr7 ( S9C Fig ) . Another interesting observation is the presence of human endogenous retroviral elements ( HERVs ) at the boundaries of both Jct1 and Jct2 , which are known to promote genome instability and induce CNV formation . [17] In Family 12 , Mother 12 ( Mat12 ) had two children with CNVs in the long arm of chr19: an ~3 . 5 Mb duplication at 19q13 . 33q13 . 41 in her daughter ( P12_dup ) , and a slightly smaller deletion ( ~3 . 352 Mb ) at the same locus in her son ( P12_del ) ( S1 Table ) . FISH analysis demonstrated an apparently balanced insertion of a segment at 19q13 . 33 into the short arm of chr19 at 19p13 in Mat12; FISH analysis also demonstrated the same insertion in P12_dup , indicating the duplication present in P12_dup was likely a recombination product of intrachromosomal maternal insertion ( S10 Fig ) . The reciprocal deletion present in P12_del was also likely a recombination product ( S11B Fig ) . High-density aCGH revealed that the apparently balanced insertion in Mat12 was not balanced—at both proximal and distal boundaries of the duplication/deletion in her two children , there were two small duplications of ~111 kb and ~77 kb in size , respectively . In addition , a small triplication ( ~33 kb ) was found embedded in the duplication near the proximal side in P12_dup ( S11A Fig ) . These additional complexities were likely accompanying events with the insertion in Mat12 that was subsequently transmitted and inherited by her two children , similar to the rearrangement events involving the PLP1 observed in BAB1381 mentioned above ( S11C Fig , refer to S11 Fig for details of proposed rearrangements in Family 12 ) .
Previously , we demonstrated that confirmatory and parental studies of CNVs by FISH analysis , especially the copy-number gains identified through CMA testing , led to the discovery of chromosomal insertions at a rate as high as 1 in ~500 individuals tested . [1] A similar high rate of 1 in ~563 individuals was independently reported . [5] Although it is now widely recognized that chromosomal insertions are not rare events , [1 , 5 , 6] the underlying mechanisms for their formation remain largely unknown . Most of the previous studies on insertions were based on relatively low resolution genome analysis by clinical arrays in combination with molecular cytogenetics , FISH , and chromosome analysis; only a few breakpoint junctions have been mapped to nucleotide resolution . [5 , 18 , 19] In this study we focused on a subset of chromosomal insertions—complex insertions with additional copy-number gain or loss at the inserted site . High-density aCGH revealed additional complexities that evaded detection by CMA testing , including small triplications embedded in duplications ( in individuals Cplex6 and P12_dup ) and small CNVs in individuals with apparently balanced insertions ( in individuals BAB1381 , Mat3 , and Mat12 ) . In addition , breakpoint junction mapping and careful examination of the junction sequences provided insights into the potential mechanisms for formation of these complex insertions , leading to the observation of distinct molecular characteristics of apparently basic complex insertions versus chromothripsis-like , chromoanasynthesis insertions . Of note , only individuals with CNVs large enough to be detected by clinical microarray , and subsequently with copy-number gains large enough to be verified as insertions by FISH , were initially identified and molecularly studied . Therefore , copy-number neutral insertions , and insertions with smaller CNVs that escaped detection by clinical array or FISH validation , were selected against inclusion in this study . We categorized individuals Cplex4 , Cplex9 , and Cplex12 as basic complex insertions ( S12 Fig ) based on the following observations: first , only one duplication was observed in these individuals , in contrast to the multiple copy-number gains observed in other individuals in this study; second , a deletion was always present at the inserting site; third , none of them were de novo events ( S1 Table ) . Breakpoint junctional sequences in these individuals showed 1–3 bp microhomology or 2–3 bp small insertions; these features represent mutational signatures of breakpoint junctions observed in structural variants potentially generated by either non-homologous end-joining ( NHEJ ) , or alternatively , microhomology-mediated end-joining ( MMEJ ) or FoSTeS/MMBIR with a single template switch . [3 , 20–23] In contrast to the individuals with basic complex insertions that were potentially generated by a number of different mechanisms , individuals Cplex5 , Cplex6 and Cplex11 showed multiple CNVs including triplications . In addition , Cplex5 and Cplex11 are de novo events ( S1 Table , inheritance mode in Cplex6 is unknown ) . Breakpoint junctions’ sequences in these individuals showed longer homology ( >4 bp ) and hundreds to thousands of base pairs of templated insertions . CNVs in these individuals resembled chromoanasynthesis events[12] , and their breakpoint junctions features are signature findings observed in structural variants generated through replicative repair based mechanism , e . g . FoSTeS/MMBIR with multiple iterative template switch events . [24 , 25] Interestingly , one of the 16 individuals identified with complex insertions initially included in this study , Cplex2 , was included and analyzed in detail in the paper that first defined the chromoanasynthesis phenomenon ( BAB3105 from Ref . 12 ) . This further strengthens our proposal that complex insertions could be part of a chromoanasynthesis event . Currently , three similar yet distinct types of chromothripsis , or chromothripsis-like events have been described , together they were referred to as ‘chromoanagenesis’ . [26] In somatic changes in the cancer genomes , chromothripsis was shown to be a catastrophic , one-step event leading to a signature pattern of frequent oscillations between unaltered and deleted copy-number states . [8] In cancer chromothripsis , most CNVs observed from genomic sequence analyses are deletions , with much less duplications resolved , and usually involves one chromosome . In contrast to the frequent copy-number loss in cancer chromothripsis , germline chromothripsis observed in individuals with neurodevelopmental abnormalities was shown to be balanced rearrangements—although several chromosomes were shattered and rejoined , the overall complex rearrangement involved almost no copy-number changes ( except for deleting or inserting short sequences at breakpoint junctions ) . [9 , 10] A recent study on unbalanced interchromosomal translocations revealed two individuals with de novo chromothripsis translocations generated through at least 18 or 33 breakpoint junctions , respectively , and both individuals only carried two large deletions ( from 800 kb to 6 . 6 Mb ) but no copy-number gains . [27] Both somatic and constitutional chromothripsis were proposed to be generated by NHEJ , given that the vast majority of the breakpoint junctions in these events showed blunt ends , 1 or 2 bp microhomology , or small insertions . [10 , 11 , 28] In contrast to the balanced germline chromothripsis involving shattering and rejoining of several chromosomes , another type of chromothripsis-like events , observed by high-density aCGH and mechanistically defined as chromoanasynthesis , was shown to involve multiple copy-number changes , particularly multiple gains of copy-number including duplications and triplications . [12] Notably , chromoanasynthesis was considered to be region-focused events . [12 , 29 , 30] In the original paper that defined the chromoanasynthesis phenomenon , all 17 individuals studied showed CNVs on the same chromosome , more specifically , 15 out 17 individuals showed CNVs confined within the distal half of the involved chromosome arms . [12] It was proposed that co-occurrence of CNVs with substantial interchromosomal exchanges would result in a non-viable offspring . [10] Here , we demonstrated that chromoanasynthesis could involve two or even three chromosomes , as we observed a templated insertion as long as 5 , 167 bp from a third chromosome in addition to the two chromosomes involved in the rearrangements in Cplex11 ( S7 Fig ) . We categorized individuals Cplex5 , Cplex6 , and Cplex11 as chromoanasynthesis events ( S12 Fig ) based on the observations that included: i ) multiple copy-number gains , including triplications , were detected , ii ) longer microhomology ( >4 bp ) observed at breakpoint junctions and iii ) long templated insertions from multiple genomic loci also present at breakpoint junctions . Similar to previously reported region-focused chromoanasynthesis events , these features are likely found in CGRs generated by iterative template switching during replicative repair based mechanisms , e . g . FoSTeS/MMBIR . [3 , 24 , 25] Note that in individuals Cplex6 and Cplex11 , some of their breakpoint junctions sequences showed blunt ends ( S5B and S7 Figs ) ; it is not uncommon to observe that a portion of the junctions in CGRs potentially generated through replication based mechanisms can show blunt ends , small insertions or short microhomology ( 1 or 2 bp ) . [15 , 31 , 32] In studies conducted in the yeast model organism , FoSTeS/MMBIR has been demonstrated to occur in the absence of microhomology ( with 0–6 bp homology at breakpoint junctions ) . [33] Although breakpoint junctions with short microhomology ( 1–3 bp ) have been observed in rearrangements proposed to be potentially generated through NHEJ , MMEJ and MMBIR in the human genome , iterative template switches are unique to the mechanism of FoSTeS/MMBIR . Therefore , it is important to consider not only microhomology length , but also the occurrence of templated insertions at junctions , and other evidence for potential iterative template switch events , in addition to whether copy-number gains ( especially triplications ) are present , when postulating potential biological mechanisms responsible for the generation of CGRs . Recent studies on DNA damage in micronuclei provided a potential further explanation for the chromothripsis and chromoanasynthesis events . Micronuclei are common outcomes of cell division defects; they are structurally similar to intact nuclei , but contain only one or a few chromosomes or chromosomal segments . [34] They could undergo defective and asynchronous DNA replication , resulting in DNA damage and extensive chromosomal fragmentation including catastrophic processes like chromothripsis; most importantly , their damaged and rearranged DNA fragments could be integrated back into the genome . [32 , 35] Rearrangements proposed to be generated through NHEJ or MMBIR have been observed in micronuclei DNA , and segments from a single chromosome were observed in the majority of the micronuclei—potentially explaining why most observed chromoanasynthesis events are chromosome or chromosome region-focused . The rare chromoanasynthesis events involving two or three chromosomes we observed in this study are potentially in accordance with the rare observation of micronuclei DNA from two chromosomes undergoing chromothripsis . [35] In this study , we also discovered that some apparently balanced insertions are actually unbalanced insertions; small deletions and duplications could be generated accompanying the inserting process . From the mechanistic aspect , it is crucial to reveal these small CNVs—a completely balanced insertion could be attributed to mechanisms like NHEJ , however , the additional CNVs , especially the copy-number gains , are more parsimoniously explained by replicative repair based mechanisms . For example , in the family with PLP1 insertion , the most parsimonious explanation for the small CNVs at both the inserting site 19q13 . 42 and missing proximal/additional distal segments accompanying the inserted fragment from Xq22 . 2 is FoSTeS/MMBIR . During the replication process , a stalled replication fork at chr19 invaded and annealed to chrX , and after replication of a genomic interval containing the entire PLP1 gene on Xq22 . 2 , the replication fork switched back to chr19q13 . 42 , however , to a more proximal locus , therefore leading to the small duplication on chr19 ( Fig 4C ) . We consider the situation in Family 12 to be similar to the PLP1 family , due to the two duplications on both boundaries of the inserting fragment potentially generated accompanying the chromosomal insertion in the mother Mat12 ( S11 Fig ) . Whereas in Family 3 , the situation may be different—unlike in the chromoanasynthesis subjects and in BAB1381 , whose CNVs are most parsimoniously explained by template switching during the replication process ( copying material from the inserting chromosomes to the inserted loci , always one direction ) , there was an exchange of genomic segments between chr9 and chr7 in Mat3 . In addition , no copy-number gain was observed in this family , and the insertions in Mat3 are mostly balanced except for the 4 kb deletion at chr7 . We propose the bi-directional , mostly balanced insertions in Mat3 may result from multiple breakages and re-joining of both chr7 and chr9 , therefore may be generated through NHEJ or MMEJ . [9 , 27] LCRs and repetitive elements are known to facilitate genomic rearrangements . [12 , 29 , 36 , 37] Enrichment of breakpoint in these repetitive sequences has been observed in nonrecurrent and complex structural changes at multiple genomic loci . [29 , 38 , 39] In the current study , we observed involvement of LCRs at breakpoint junctions in individuals Cplex5 and Cplex6 ( Table 1 ) , and HERV elements at breakpoint junctions in Family 3 and Cplex6 ( S9C Fig ) . In addition , we observed involvement of other repetitive sequence , e . g . SINEs ( short interspersed nuclear elements ) at junctions in Cplex9 , Cplex6 , BAB1381 and in Family 12 , also LINEs ( long interspersed nuclear elements ) at junctions in Cplex4 , Cplex12 , Cplex5 , Cplex11 and BAB1381 ( Table 1 ) . These repeat and repetitive sequences may stimulate genomic instability and potentially assist replicative repair catalyzed genomic rearrangements facilitating template switching and the generation of the nonrecurrent and complex insertion events . [3 , 29 , 40] In summary , from studies of complex chromosomal insertions , we observed that chromoanasynthesis could occur beyond a confined chromosomal region and involve two or three chromosomes . We observed microhomologies and templated insertions at the breakpoint junctions , resembling the breakpoint junction signatures found in CGRs generated through replication-based mechanism ( s ) and iterative template switches: FoSTeS/MMBIR . [3] We propose that DNA replicative repair mechanisms can potentially result in interchromosomal complex insertions , and cause a significant fraction of apparently balanced insertions; especially those harboring small flanking CNVs .
Sixteen individuals with complex chromosome insertions were identified in the CMA laboratory at Baylor Genetics among the ~56 , 000 individuals tested from 2007 to 2014 . This study was approved by the Institutional Review Board for Human Subject Research at Baylor College of Medicine ( IRB H-25466 ) . Informed consent was obtained prior to collecting identifiable DNA samples ( BAB1379 , BAB1380 , BAB1381 , P3 , Mat3 , P12_del , P12_dup and Mat12 ) . The remaining DNA samples were de-identified for breakpoint and mechanistic studies ( named Cplex1 , Cplex2 , Cplex3 , etc ) . Custom designed BCM OLIGO V6 . 5 , V7 , V8 , V9 or V10 oligonucleotide arrays were performed as previously described . [41 , 42] Arrays were designed to specifically interrogate clinically significant regions with an average resolution of 30 kb between probes . Interphase and metaphase FISH were performed to confirm the CMA findings and tested using available parental samples . [1] To further characterize the CNVs identified by CMA and FISH involving complex insertions , we designed several 4X 180K oligonucleotide arrays with ~200 bp per probe spacing from Agilent Technologies ( AMADID 073188 , 073189 , 076797 , 079204 , 071585 , 024241 and 015482 ) . Hybridization controls were gender matched ( Individual NA10851 as male control and Individual NA15510 as female control ) . Scanned array images were processed using Agilent Feature Extraction software ( version 10 ) and extracted files were analyzed using Agilent Genomic Workbench ( version 7 . 0 . 4 . 0 ) . Array designs and sequence alignment for breakpoint analysis were based on the February 2009 genome build ( GRCh37/hg19 assembly ) . To further confirm the CNVs identified by high-density arrays and map the breakpoint junctions , primers flanking the predicted breakpoints were designed and long-range PCRs were conducted using TaKaRa LA Taq according to the manufacturer’s protocol ( TaKaRa Bio Company , Cat . No . RR002 ) as previously described . [29] PCR products were prepared for sequencing using ExoSAP-IT ( Affymetrix , Cat . No . 78201 ) according to the manufacturer’s protocol or gel extracted and purified with the Zymoclean Gel DNA Recovery kit ( Zymo Research , Cat . No . D4001 ) . Purified PCR products were then sequenced by Sanger di-deoxynucleotide sequencing ( BCM Sequencing Core , Houston , TX , USA ) . To elucidate the insertion site in individual Mat3 , the APAgene GOLD genomic walking kit was used according to the company’s protocol ( BIO S&T , Cat . No . BT901-RT ) . Generally , this kit enables isolation of unknown sequences which flank known sequences . Three rounds of nested PCR with degenerate random tagging primers provided by the kit were performed , and the end PCR products were cloned into a TA vector ( pGEM-T Easy Vector Systems , Promega , Cat . No . A1360 ) and were further subjected to Sanger sequencing . | By traditional cytogenetic techniques , the incidence of microscopically visible chromosomal insertions was estimated to be 1 in 80 , 000 live births . More recently , by aCGH in conjunction with FISH confirmation of the aCGH findings , insertion events were demonstrated to occur much more frequently ( 1 in ~500 individuals tested ) . Although frequently detected , little is known about the molecular mechanisms of their formation . In this study , we identified 16 individuals with complex chromosomal insertions among 56 , 000 individuals tested at Baylor Genetics using clinical microarray analysis ( CMA ) and FISH . Custom high-density aCGH was performed on 10 individuals with available DNA , and breakpoint junctions were fine-mapped at nucleotide resolution by long-range PCR and DNA sequencing in 6 individuals to glean insights into potential mechanisms of formation . In addition , we analyzed 5 families with apparently balanced insertion in one parent detected by FISH analysis and found that 3 parents had additional small copy-number variants ( CNVs ) at one or both sides of the inserting fragments as well as at the inserted sites . We propose that replicative repair can result in interchromosomal complex insertions generated through chromothripsis-like chromoanasynthesis involving two or three chromosomes , and cause a significant fraction of apparently balanced insertions harboring small flanking CNVs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"probe",
"techniques",
"dna",
"replication",
"molecular",
"biology",
"techniques",
"translocations",
"dna",
"duplications",
"research",
"and",
"analysis",
"methods",
"genome",
"complexity",
"probe",
"hybridization",
"chromosome",
"biology",
"chromosomal",
"aberrations",
"molecular",
"biology",
"biochemistry",
"fluorescent",
"in",
"situ",
"hybridization",
"cell",
"biology",
"nucleic",
"acids",
"cytogenetic",
"techniques",
"genetics",
"deletions",
"biology",
"and",
"life",
"sciences",
"genomics",
"computational",
"biology",
"chromosomes"
] | 2016 | Mechanisms for Complex Chromosomal Insertions |
Traditionally microorganisms were considered to be autonomous organisms that could be studied in isolation . However , over the last decades cell-to-cell communication has been found to be ubiquitous . By secreting molecular signals in the extracellular environment microorganisms can indirectly assess the cell density and respond in accordance . In one of the best-studied microorganisms , Bacillus subtilis , the differentiation processes into a number of distinct cell types have been shown to depend on cell-to-cell communication . One of these cell types is the spore . Spores are metabolically inactive cells that are highly resistant against environmental stress . The onset of sporulation is dependent on cell-to-cell communication , as well as on a number of other environmental cues . By using individual-based simulations we examine when cell-to-cell communication that is involved in the onset of sporulation can evolve . We show that it evolves when three basic premises are satisfied . First , the population of cells has to affect the nutrient conditions . Second , there should be a time-lag between the moment that a cell decides to sporulate and the moment that it turns into a mature spore . Third , there has to be environmental variation . Cell-to-cell communication is a strategy to cope with environmental variation , by allowing cells to predict future environmental conditions . As a consequence , cells can anticipate environmental stress by initiating sporulation . Furthermore , signal production could be considered a cooperative trait and therefore evolves when it is not too costly to produce signal and when there are recurrent colony bottlenecks , which facilitate assortment . Finally , we also show that cell-to-cell communication can drive ecological diversification . Different ecotypes can evolve and be maintained due to frequency-dependent selection .
Complex systems in biology often come about through the communication of their parts , such as pheromone communication in insect societies and language in humans . Communication has been found to be ubiquitous in microorganisms as well [1]–[4] . Due to self-produced molecular signals that are secreted in the environment , cells can monitor the population density , which can quantitatively affect a cell's gene expression or trigger a differentiation process . In 1994 , Fuqua and colleagues were the first to characterize this form of cell-to-cell communication as quorum-sensing signaling [5] . Quorum-sensing signaling has been shown to regulate a multitude of bacterial processes , such as extracellular enzyme production , antibiotic production and biofilm formation [6]–[11] . In one of the best-studied microorganisms , Bacillus subtilis , the differentiation of a number of cell types has been shown to depend on cell-to-cell communication [12]–[14] . These cell types emerge during the developmental process of biofilm formation and are presumably needed to survive the harsh environmental conditions that are present in the soil [10] , [15] , [16] . The most remarkable survival strategy among these cell types is that of the spore [17] , [18] . A spore is a metabolically inactive cell that compartmentalized its DNA together with some essential proteins to survive starvation or other environmental stressors [18] , [19] . Spore formation is an energy-expensive process that can take 6 to 8 hours and involves the expression of hundreds of genes [19] , [20] . The initiation of sporulation is primarily dependent on the activation of a single transcription factor called Spo0A [14] , [21]–[24] . When the level of activated Spo0A is sufficiently high , the sporulation process will be initiated [25]–[28] . The level of activated Spo0A is indirectly affected by a number of environmental and physiological cues , of which some are self-produced quorum-sensing signals [13] , [22] , [29] . These signals are assumed to accumulate in the environment and thereby give an indication of the cell density . As a consequence , the fraction of cells that initiate sporulation is higher for higher cell densities [29]–[33] . Even though these quorum-sensing signals affect the proportion of cells that initiate sporulation , they themselves are not sufficient for initiating sporulation since starvation is absolutely required [34]–[36] . Bischofs and colleagues ( 2009 ) mathematically modeled the regulatory mechanisms that integrate the quorum-sensing signals with other environmental cues , including those that are indicative of starvation [36] . They showed that the quorum-sensing signals allow for a density-dependent normalization of certain environmental cues . For example , when a cell can sense the amount of nutrients that are left in the environment , quorum-sensing signaling makes it possible to estimate the amount of nutrients that are left per cell . They concluded that these density-dependent normalizations might be adaptive for cellular decision-making , such as determining when to initiate sporulation ( see also [34] ) . However , despite the detailed knowledge of the regulatory mechanisms that underlie the sporulation process , little is known about their evolutionary origin . Why does cell-to-cell communication evolve and under which ecological and developmental conditions is it selected for ? Here we examine , by using individual-based simulations , how three conditions , which inevitably relate to sporulation [15] , [19] , [20] , [34] , [37] , [38] , affect the evolution of cell-to-cell communication: environmental variation in nutrient conditions , costs of sporulation and time expenditure of sporulation . Even though our model is inspired by sporulation in B . subtilis , it is aimed to be conceptual and therefore does not include mechanistic details . The model is made such that it allows for the evolution of various developmental strategies , in which a cell's sensitivity and response to environmental cues can evolve . Throughout the paper we discuss different versions of our model , which gradually increase in complexity . First we study the evolution of cell-to-cell communication under clonally-growing colonies . Next we allow for within colony-variation by initiating colonies with multiple individuals . Under these conditions multiple ecotypes evolve that transiently coexist over time due to negative frequency-dependent selection . Finally , we examine the evolution of cell-to-cell communication when signal production is costly . Under these conditions cooperative dilemmas emerge naturally and we find that different ecotypes evolve , which use different communicative strategies to time the onset of sporulation . The evolutionary significance of these strategies can only be understood by considering their ecological context .
We assume that cells are scattered throughout the soil . Only in a few locations these cells can grow and form colonies , because only in these areas there are nutrients available to do so . During colony growth cells consume nutrients in order to perform cell division and cell differentiation . A cell can differentiate into two cell types—a signal-producing cell or a spore—or it could remain undifferentiated . Eventually , all the nutrients will be depleted and a colony enters a starvation period . This period can only be survived by the spores . It is therefore crucial for a cell to initiate sporulation on time ( i . e . when the nutrients that are needed to complete the sporulation process are still available ) . To decide when to initiate sporulation a cell could make use of two environmental cues: the nutrient concentration and the amount of quorum-sensing signal . The spores that eventually survive the starvation period migrate and germinate in new nutrient rich areas , where they form new colonies . Over evolutionary time , a cell's responsiveness to the environmental cues can evolve and thereby the timing of sporulation can evolve as well . We examine under which ecological and developmental conditions there is selection for cells that use quorum-sensing signaling to time the onset of sporulation . The system is studied by using individual-based simulations , which we describe in the following paragraphs . We assume that the population of cells is divided into subpopulations , each representing a colony ( i . e . biofilm or pellicle ) . Each colony is established by individuals . A colony is said to grow clonally when it is established by only one individual ( ) . At the onset of colony growth there is a single nutrient input , which for each colony is taken from a normal distribution that is given by . Thus , the nutrient could be different for each colony . After receiving the nutrient input colonies are allowed to grow for a fixed number of time steps ( ) ; during this period cells consume nutrients in order to perform cell division and differentiation . At the end of a nutritional cycle all individuals ( cells and spores ) enter migration . The nutritional cycles of all colonies are synchronized such that the individuals from all colonies enter migration at the same time , forming a single migratory pool ( see figure 1 ) . Since migration occurs passively , we assume that all individuals have the same chance to establish a new colony . Thus , new colonies are established by choosing , for each colony separately , random individuals from the migratory pool . After this , the new colonies simultaneously start the next nutritional cycle . Within a nutritional cycle three different cellular processes can occur at any time step ( for each cell in the colony ) . First , a cell gets the opportunity to differentiate . A cell can differentiate into two different cell types—a signal-producing cell or a spore—or it could remain undifferentiated . A signal-producing cell secretes a fixed amount of signal in the environment . The more cells that produce signal , the higher the amount of extracellular signal . At the same time , the signal is degraded with a fixed rate . Thus , the amount of signal changes over time depending on the number of cells that are producing it . A cell could also initiate sporulation . Sporulation is an irreversible process that takes a fixed number of time steps ( ) and during which a fixed amount of nutrients is consumed ( ) , which is needed for making the spore . Thus , a sporulating cell consumes nutrients per time step . When there is an insufficient amount of nutrients in the environment , the sporulation process cannot be completed; in this case a cell inevitably dies . After completing the sporulation process , a mature and resistant spore is formed . A spore cannot divide , but has a much lower death rate than a cell . A spore germinates at the onset of a new nutritional cycle . Since sporulation requires time steps , a cell can be in one , out of , phenotypic states . It can be an undifferentiated cell , a signal producing cell or a sporulating cell , of which the latter is subsequently composed of states that indicate the number of time steps a cell has been sporulating ( ) . At the final time step of sporulation ( ) a cell turns into a spore . The cell's decision to differentiate into a signal-producing cell or spore depends in our model on two environmental cues—the amount of nutrients and signal—and on a cell's genotype ( which we describe later ) . The second cellular process that a cell can undergo , after having had the opportunity to differentiate , is division . All cells , excluding spores , have a certain chance of dividing . This chance is dependent on the amount of nutrients that are present in the environment ( for details see equation S1 ) . The more nutrients that are present in the environment , the greater the chance of cell division , with a maximum chance of . During each cell division a fixed amount of nutrients ( ) is consumed . At each cell division there is a certain probability that the dividing cell incurs a mutation ( the mutation process is described later ) . The third and last cellular process that can occur at any particular time step is that of cell death . Both cells and spores have a fixed chance of dying , which is independent of the nutrient concentration . The death rate of a spore is much lower than that of a cell ( ) . Hence , it is better to be a cell when nutrients are plentiful , because the chance of having cell division outweighs the chance of having cell death . On the contrary , when the nutrients are depleted , it is better to be a spore because spores have a smaller chance of dying than cells . The fitness of a genotype therefore depends on the timing of sporulation . When a genotype sporulates too early—at a nutrient concentration that is too high—it loses reproductive potential , since not all the nutrients are utilized . When a genotype sporulates too late—at a nutrient concentration that is too low—it has an increased risk of dying , especially when , due to nutrient scarcity , the sporulation process cannot be completed . A crucial part of the model is the cell differentiation process . We aim to model it such that various developmental strategies can evolve . This requires to have sufficient degrees of freedom . On the other hand , we want to restrict the number of evolvable variables , in order to keep the model simple and tractable . The combination of these requirements resulted in a cell differentiation process that could be described by two Boolean decision-making steps , which are affected by the amount of nutrients and signal . The cell should decide to initiate sporulation or not and when it does not sporulate , a cell should decide if it wants to produce signal or not . These two decisions can be expressed by the following two inequalities ( see figure 2 ) : ( 1a ) ( 1b ) Inequality 1a shows when a cell initiates sporulation and inequality 1b shows when a cell initiates signal production . We assume that the decision to initiate sporulation is dominant over the decision to produce signal . Thus when both inequalities hold , only the sporulation process is initiated . The left hand side of each inequality contains the environmental cues: the amount of nutrients ( ) and the amount of extracellular signal ( ) . Since nutrients are consumed and signal can be produced and degraded over time , the values of these environmental cues change during colony growth . The effect of an environmental cue on the differentiation process depends on what we call the connection weight , ; here is the environmental cue ( 1 is the amount of nutrients and 2 is the amount of signal in the environment ) that is affecting differentiation process ( 1 is sporulation and 2 is signal production ) . For example , determines how the amount of nutrients affects the initiation of sporulation . When a connection weight is positive , its corresponding environmental cue stimulates the differentiation process . When the connection weight is negative , the environmental cue inhibits the differentiation process . The absolute value of a connection weight shows the impact that a certain environmental cue has on the differentiation process . The right hand side of both inequalities is the activation threshold , ; here is the differentiation process to which the activation threshold belongs ( 1 is sporulation and 2 is signal production ) . The activation threshold shows how much stimulus from the environmental cues is required before the differentiation process is initiated . For example , when is positive a cell only sporulates when the stimulus from the nutrients ( ) plus the stimulus from the signal ( ) is bigger than the activation threshold ( ) . On the contrary , when is negative a cell sporulates by default ( when ) and sporulation can only be prevented if the environmental cues inhibit the sporulation process ( i . e . negative connection weights ) . The activation thresholds could be viewed as a normalization of the connection weights . Namely , one could divide both sides of inequality 1a and 1b by the absolute values of , respectively , and , without altering the behavior of a genotype . Therefore the model could be simplified by fixing the activation thresholds ( i . e . preventing mutations to occur in the activation thresholds ) , as long as it does not affect the strategies that can evolve . In the first two sections of the results we applied this simplification to the model and only allowed the connection weights to mutate . To show that this simplification did not affect the evolutionary outcome of the model we performed all simulations under non-simplified conditions and show the results in the supplementary information ( figure S3 ) . In the last section we did not fix the activation thresholds , because when signal production is assumed to be costly , the evolutionary outcome would be constrained by fixing the activation thresholds . We call the collection of connection weights ( ) and activation thresholds ( ) the genotype of an individual . In essence , the genotype describes how a cell responds to each combination of environmental cues . When a cell division occurs each of the genotypic variables ( and ) has a certain chance to mutate ( ) . When a mutation occurs , a small value taken from the normal distribution is added to the genotypic variable . Every mutation is taken independently from the same normal distribution , irrespective of the genotypic variable that mutates . All evolutionary simulations are initiated with the same monomorphic population of cells that do not produce signal and are not sensitive to it ( ) . In addition , the initial cells are assumed to sporulate , to prevent the population from going extinct . The initial cells sporulate at a nutrient concentration of 500 ( and ; all input variables that are perceived by the cells are divided by 1000 as normalization , which is done consistently throughout the paper ) . Similar results would however be obtained if sporulation would occur at another nutrient concentration , as long as the initial population does not go extinct in the first growth cycle . By assuming that both and are negative , we assume that nutrients inhibit the sporulation process and that when this inhibition is too weak ( e . g . when ) a cell initiates sporulation . Thus , we are not examining the evolution of sporulation , but the evolution of cell-to-cell communication as a mechanism to time the onset of sporulation .
In this section we examine the evolution of cell-to-cell communication under the assumption that colonies grow clonally , meaning that colonies are initiated by a single individual ( ) . Genetic variation can only arise in these colonies via mutations . Moreover , for simplicity as explained before , we also assume that only the connection weights ( ) can mutate ( similar results are however obtained when the activation thresholds are allowed to mutate as well; see figure S3 ) . Under these conditions , the timing of sporulation depends on and and the differentiation into a signal-producing cell solely depends on and ( the activation thresholds , , are fixed over evolutionary time ) . To evolve cell-to-cell communication a cell should acquire two properties over evolutionary time . First , a cell should produce signal . Thus , before initiating sporulation a cell has to differentiate into a signal-producing cell . Second , a cell should be sensitive to the signal ( ) , meaning that the nutrient concentration at which a cell initiates sporulation has to depend on the amount of signal . Irrespectively of the order in which these properties evolve , when both are present there is cell-to-cell communication . To examine if both properties can evolve in our model , we ran individual-based simulations that were initiated with a monomorphic population of cells that did not produce signal and were not sensitive to the signal ( ) . Figure 3A shows two independent evolutionary trajectories projected on an adaptive landscape ( for more replicates see figure S1 ) . The adaptive landscape is constructed by showing for each possible genotype—meaning each combination of and —the average colony size that is obtained at the end of a nutritional cycle . When solely examining the adaptive landscape , one expects that cell-to-cell communication would evolve , because the best-performing genotypes that are signal-sensitive ( ) have a higher fitness than those that are signal-insensitive ( ) . The two evolutionary trajectories that are plotted on the adaptive landscape are called run 1 and run 2 ( both runs were performed under the same parameter settings ) . In both runs cell-to-cell communication evolved , which means that both signal-production and signal-sensitivity evolved . The evolutionary trajectories of figure 3A and S1 closely match the adaptive landscape and hence the adaptive landscape can be used to predict the outcome of evolution . The adaptive landscape only shows the selective advantage of cell-to-cell communication for and since nothing interesting happens outside this quadrant . In other words , nutrients are expected to inhibit sporulation ( i . e . a cell only sporulates when there is nutrient scarcity ) , while signal is expected to stimulate sporulation ( i . e . a cell sporulates earlier when it occurs in a bigger population ) . A limitation of the adaptive landscape of figure 3A is that it does not show the other two connection weights , and . and determine when a cell differentiates into a signal-producing cell ( see figure 2 ) . Signal production is , next to signal-sensitivity , essential for the evolution of cell-to-cell communication . To examine how signal production evolved we plotted the values of all connection weights ( corresponding to the most-abundant genotypes ) , of run 1 , along a time-axis ( see figure 3B ) . Figure 3B shows that signal production evolves after about 20 . 000 time steps ( becomes positive; as indicated by the green arrow ) . About 40 . 000 time steps later signal-sensitivity evolves as well ( becomes positive; as indicated by the blue arrow ) . In other words , signal production emerges before the occurrence of signal-sensitivity . Hence there was no selective advantage for signal production at the moment it evolved . Signal production evolved because a neutral mutation in hitchhiked along with a beneficial mutation in . Genetic hitchhiking is relatively prevalent , because there is no genetic recombination . In addition , there are no costs for signal production in this version of the model . Thus , cell-to-cell communication evolves by the sequential evolution of signal production and signal-sensitivity . The question we are interested in though , is why cell-to-cell communication evolved at all . By sensing signal a cell can assess the colony size at the onset of sporulation . This estimate gives an indication of the amount of nutrients that will be consumed by the colony during sporulation . As explained before , a cell should turn into a spore when the chance of having cell death exceeds that of cell division , which is associated with a critical nutrient concentration ( for details see equation S2 ) . Since sporulation requires time , a cell has to anticipate or predict if the nutrient concentration at the end of sporulation matches this critical nutrient concentration . To make this prediction it is necessary to assess the amount of nutrients that will be consumed during sporulation . Since the total amount of nutrient consumption depends on the number of cells within a colony , it is advantageous for a cell to sense quorum-sensing signals . When the colony is big , a high amount of nutrients will be consumed during sporulation due to which a cell should initiate sporulation relatively early ( i . e . at a high nutrient concentration ) . On the contrary , when the colony is small , a small amount of nutrients will be consumed and therefore a cell should initiate sporulation relatively late ( i . e . at a low nutrient concentration ) . Thus , cell-to-cell communication allows a cell to predict the total amount of nutrient consumption during sporulation and , thereby , a cell can anticipate future environmental changes . There are three requirements that should be satisfied for cell-to-cell communication to evolve ( corresponding to the parameter values in our model; see figure 4 ) : ( i ) the colony size should affect the nutrient concentration during sporulation by , for example , nutrient consumption ( ) ; ( ii ) there should be a time-lag between the moment that a cell decides to sporulate and the moment that it turns into a mature spore ( ) ; and ( iii ) there should be environmental variation ( ) . High values of , and ( e . g . , and ) can result in a fitness advantage for cells that sense quorum-sensing signals over those that do not ( figure 4 ) . The first requirement for the evolution of cell-to-cell communication is that the colony size should affect the nutrient concentration ( figure 4A ) . For example , when each cell consumes a fixed amount of nutrients during sporulation ( ) , the total nutrient consumption depends on the colony size . When there is no nutrient consumption during sporulation ( ) the optimal time at which to initiate sporulation does not depend on the colony size and hence cell-to-cell communication does not evolve . Second , cell-to-cell communication only evolves when there is a time-lag between the moment that a cell decides to sporulate and the moment that it turns into a spore ( figure 4B ) . In other words , sporulation should require time . When sporulation does not require time , there is no need to assess the nutrient consumption since a cell could turn into a spore instantaneously . Thus , cell-to-cell communication only evolves when . The third and last requirement for the evolution of cell-to-cell communication is the presence of environmental variation ( figure 4 ) . When there is no variation ( ) , the amount of nutrients at the onset of a nutrient cycle is always the same . As a consequence , the changes in the nutrient concentration over time correlate with those of the colony size , since all colonies are initiated with the same number of cells , which reproduce at the same rate . Under these conditions , the nutrient concentration could be used as an accurate indication of the colony size , which makes the use of quorum-sensing signals superfluous , since these give an indication of the colony size as well . Only when the correlation between the nutrient concentration and colony size is relatively weak , the amount of signal could be used as a unique indication of the colony size . For this reason , there is stronger selection for cell-to-cell communication for higher levels of . Alternative conditions that weaken the correlation between the colony size and nutrient concentration can have a similar effect . For example , one could vary the initial colony sizes; colonies would still be clonal but different colonies would be initiated by different numbers of cells ( see figure S8 ) . In most laboratory experiments sporulation is studied in isogenic populations . However , it is plausible that multiple genotypes can co-occur in a single colony [39] . In this section we examine how the developmental mechanisms that determine the onset of sporulation evolve when multiple genotypes can initiate a single colony ( ) . This is done for the same conditions as those described in the previous section ( i . e . only the connection weights , , are allowed to mutate; see figure S3 for simulations in which also the activation thresholds could mutate ) . In figure 3C the evolutionary trajectory of a single run is shown on the adaptive landscape . Figure 3D shows , for the same evolutionary run , the connection weights of the most-abundant genotypes along a time-axis ( for more replicates see figure S2 ) . In contrast to the previous section , there is a bifurcation event during the evolutionary process that results in two coexisting ecotypes ( an ecotype is a cluster of genotypes that is adapted to specific ecological condition ) . One of these ecotypes eventually goes extinct ( see figure 3D and S2 ) . Both ecotypes produce quorum-sensing signal and are sensitive to it . The ecotypes only differ in their responsiveness towards the nutritional conditions in the environment ( ) . In one ecotype the value of is lower than in the other , meaning that the nutrients more strongly inhibit the sporulation process ( see figure 3D and S2 ) . This ecotype is therefore called the late sporulating ecotype ( i . e . sporulation is initiated at a low nutrient concentration ) , while the other one is called the early sporulating ecotype ( i . e . sporulation is initiated at a high nutrient concentration ) . How can the late and early sporulating ecotypes stably coexist ? In the absence of cell-to-cell communication , a genotype can only efficiently make use of the available nutrients for a limited range of nutrient inputs ( i . e . nutrient concentration at the onset of a nutritional cycle; see figure S4 and S5B ) . When the nutrient input is higher than this particular range , a genotype would sporulate too late and when it is lower than this range a genotype would sporulate too early ( see figure S4 ) . When a genotype sporulates too early , not all the nutrients will be consumed . The leftovers can be used by other genotypes that sporulate slightly later and co-occur in the same colony . The late sporulating genotypes , in turn , cannot efficiently make use of the nutrients at high nutrient inputs , because they initiate sporulation too late . As a consequence , there is frequency-dependent selection in which the late sporulating ecotype has a selective advantage when the early sporulating ecotype is abundant and vice versa ( see figure S6 ) . Figure 3D shows that the early sporulating ecotype evolves first and later is accompanied by the late sporulating ecotype . Over evolutionary time both the early and late sporulating ecotypes become more sensitive to the quorum-sensing signal ( increase in ) and thereby evolve cell-to-cell communication ( figure 3D ) . In other words , both ecotypes evolve the ability to adjust the timing of sporulation to the nutrient input . This increases the range of nutrient inputs at which an ecotype could efficiently make use of the nutrients ( see figure S5C ) . As a consequence , there is an increasing overlap in the range of nutrient inputs at which both ecotypes grow efficiently , hence strengthening the competition between them . Ultimately , only a single ecotype survives ( see figure 3D and S2 ) . This ecotype is a generalist , since it grows efficiently at most nutrient inputs due to the evolved cell-to-cell communication . Thus , over evolutionary time , the evolved specialists—the early and late sporulating ecotypes—are replaced by a generalist—a signaling ecotype—that can grow efficiently at most nutrient inputs . Not surprisingly , when there is no environmental variation ( ) , a bifurcation event cannot occur . In that case only a single ecotype evolves that outcompetes all others ( see figure S7 ) . Branching is most likely to occur for high levels of ( see figure S7 ) ; the same conditions that select for cell-to-cell communication ( see figure 3 and 4 ) . Another condition under which a bifurcation event cannot occur is clonal growth , since it hampers the presence of within-colony variation . Within-colony variation allows for competition at the cellular-level and hence for the coexistence of multiple ecotypes . However , allowing for within-colony variation can also result in a conflict between the genotypes that are selected for at the colony-level and those that are selected for at the cellular-level . In particular , when signal production is costly conflicts are expected , since cells that do not produce the costly signal have a fitness advantage at the cellular-level but undermine the performance of the colony . In the next section we examine whether cell-to-cell communication evolves when signal production is costly . In this section we examine whether cell-to-cell communication can still evolve when signal production is costly . We assume that a signal-producing cell has a reduced chance of dividing by subtracting a fixed value ( ) from the chance of having cell division ( see equation S3 ) . In contrast to the previous sections , all genotypic variables can mutate , to allow for a wider variety of communicative strategies . In this section we focus on a single representative evolutionary run ( for more replicates see figure S9 ) . Figure 5 shows the outcome of this evolutionary run , by using a phenogram . The phenogram shows the dissimilarity between genotypes in a population that evolved for 550 . 000 time steps . The genotypes are named by letter-codes , which are ranked in alphabetic order and represent abundance , with genotype ‘AA’ being the most abundant and genotype ‘CH’ the least . Besides the letter-code , every genotype is connected to a small graph , which shows its phenotype for a range of environmental conditions . The population consists of multiple communicative strategies that cluster together . The three most-abundant genotypes partly reflect these clusters and are shown on the left side of the phenogram . Since , the phenogram does not show evolutionary descendance , the evolutionary lineages of the three most-abundant genotypes were used to construct an evolutionary tree . This tree is shown in figure 6 . Hereafter , the phenotypes of the three most-abundant genotypes are called phenotype 1 , 2 and 3; corresponding to the order in which they appear in figure 6 . All three phenotypes produce quorum-sensing signal for a range of parameter conditions ( shown by the green areas in figure 6 ) . Phenotype 2 produces quorum-sensing signal for all environmental conditions , except for those at which it sporulates . Since signal production is costly this phenotype is exploited by phenotype 1 and 3 , which lack signal production for respectively high and low nutrient concentrations . As a consequence , phenotype 2 is always selected against at the cellular-level , irrespective of the population composition at the onset of a nutritional cycle . However , phenotype 2 is maintained in the population due to selection at the colony-level , in which the colonies that contain phenotype 2 often have a selective advantage over those that do not contain phenotype 2 ( for details see table S1 ) . This selective advantage results from the improved timing of sporulation . Thus , the selection pressures at the colony-level outweigh those at the individual-level . Since the other two phenotypes exploit phenotype 2 for different environmental conditions , they occupy different niches . Figure 7 shows the selection pressures that act on each phenotype , given the frequency at which each phenotype occurs in the population ( frequency over all colonies ) . The fitness measurements include the selection processes at the cellular- and colony-level . All phenotypes have a selective advantage when they are present in a low overall frequency . Thus , negative frequency-dependent selection is responsible for the stable coexistence of the three phenotypes . Since the three phenotypes are subject to a continuing process of evolution , it is unlikely that these specific phenotypes would coexist forever . Frequency-dependent selection does however assure the coexistence of multiple ecotypes , as shown by figure 5 and S9 . It is important to notice that the evolutionary simulation shown by figures 5 , 6 and 7 assumes relatively low costs for signal production and a small bottleneck size . The costs of signal production are 2% of the maximal growth rate ( ) , which means that a signal-producing cell has a 2% smaller chance to divide than an undifferentiated cell under the optimal growth conditions . The bottleneck size is given by the number of individuals that initiate a single colony ( ) . Smaller bottleneck sizes facilitate assortment , because signal-producing cells are more likely to end up in a colony that only contains signal-producers . As a consequence , signal-producing cells are less likely to be exploited by cells that lack signal production . Figure 8 shows how the evolution of cell-to-cell communication depends on and , by showing the average amount of signal that is present in a population that evolved for 550 . 000 time steps . As expected , cell-to-cell communication is more likely to evolve for smaller signal costs and stronger population bottlenecks . In conclusion , when signal production is costly , cell-to-cell communication can still evolve . However , signal-producing cells can be exploited by cells that lack signal production . This ultimately results in the evolution of ecological diversity , in which multiple ecotypes can coexist . Even though it is to be expected that signal production costs result in cheating ( i . e . cells that do not produce signal ) , it is less intuitive that three ecotypes would evolve , including one that cheats for high nutrient inputs and another that cheats for low nutrient inputs . This coexistence is facilitated by negative frequency-dependent selection , which results from the selection processes at the cellular- and colony-level . Cell-to-cell communication only emerges in our simulations for relatively low costs of signal production and in the presence of population bottlenecks .
We demonstrated that cell-to-cell communication can evolve to regulate the timing of sporulation . The evolution of cell-to-cell communication requires both the evolution of signal production and signal-sensitivity . By sensing quorum-sensing signals a cell can predict future environmental conditions and thereby anticipate a starvation period by initiating sporulation . To predict the environmental conditions a cell has to assess the rate of nutrient consumption , which depends on the colony size . Our model shows that three conditions , which inevitably relate to sporulation , are sufficient to explain the evolution of cell-to-cell communication: ( i ) the population size has to affect the nutrient concentration ( ) ; ( ii ) a cell has to predict future environmental conditions ( ; see also [40]–[42] ) ; and ( iii ) there has to be environmental variation ( ) . Irrespectively of how these conditions come about , when all three are satisfied and signal production is not too costly , cell-to-cell communication evolves . It is not our claim that these conditions are strictly necessary , but rather that they are sufficient for the evolution of cell-to-cell communication . In nature , the requirements for the evolution of cell-to-cell communication in sporulating bacteria might be less stringent , since additional advantages , besides the timing of cell differentiation , can facilitate the evolution of cell-to-cell communication ( e . g . colony-level properties; [2] ) . In contrast to previous models on the evolution of cell-to-cell communication [43]–[46] , our model shows that cell-to-cell communication can evolve as a mechanism to evaluate other environmental cues [34] , [36]: neither the absolute signal concentration nor the absolute nutrient concentration determine the onset of sporulation . To understand when cell-to-cell communication evolves one has to understand how the information that results from quorum-sensing signaling is integrated with that of other environmental cues [47]–[49] . Moreover , we have demonstrated that cell-to-cell communication can even evolve when there is genetic variation within the colony and , in addition , when signal production is costly . Models on sporulation ( or other persistence phenotypes ) often exclude cell-to-cell communication as a mechanism to regulate sporulation [27] , [50] , [51] . This is because sporulation is mostly studied as a bet-hedging strategy: only a small fraction of genetically-identical cells sporulates under the same environmental conditions [28] , [40] . Bet-hedging is a risk-spreading strategy that ensures the survival of a colony when there are severe and sudden environment changes [52] , [53] . In our model a bet-hedging strategy cannot evolve , because cells always perceive accurate environmental information and lack developmental noise . Furthermore , bet-hedging is only beneficial when environmental changes are unpredictable [50] , [54] . In our model , environmental changes might only become unpredictable when a cell is surrounded by different ecotypes , which differ in the amount of signal production and the timing of sporulation . It might therefore be interesting to extend the model , in order to examine how the evolution of bet-hedging affects that of cell-to-cell communication . In our model , cell-to-cell communication represents a form of phenotypic plasticity , because it allows a cell to adjust the timing of sporulation in response to environmental changes [55] . Without cell-to-cell communication a cell can only grow efficiently for a limited range of nutrient inputs ( figure S5 ) . In that case , multiple ecotypes evolve that specialize on distinct ecological niches ( e . g . the late and early sporulating ecotypes that evolved at the onset of our simulations , see figure 3C–D ) . However , by evolving cell-to-cell communication the range of nutrient inputs at which a cell grows efficiently increases . This ultimately results in competitive exclusion: the specialized ecotypes ( i . e . narrow niche width ) —such as the late and early sporulating ecotypes—are replaced by a single generalist ( i . e . broad niche width ) that can grow efficiently under most environment conditions due to cell-to-cell communication [56]–[58] . In our model phenotypic plasticity is a colony-level property , instead of a cellular property , since cells cannot respond to changes in environmental conditions without cooperation [59]: the amount of signal only gives an accurate indication of the colony size when all cells ( or a constant fraction ) produce quorum-sensing signals . The evolution of cell-to-cell communication therefore entails a cooperative dilemma ( given that signal production is costly; [4] , [60]–[62] ) . Cells that do not produce signal ( i . e . public good ) have an advantage over those that do , but at the same time they undermine the colony performance ( see also [4] , [63]–[66] ) . The cells that do not produce signal could therefore be called ‘cheaters’ , while signal-producing cells are ‘cooperators’ . In our model , cheaters and cooperators evolved and stably coexisted due to frequency-dependent selection [43] , [46] , [67]–[71] . They have different communicative strategies [72] and therefore occupy distinct complementary niches ( see figure 5 and 6 ) . That is , the cheaters lack signal production for different subsets of environmental conditions . This emphasizes the importance of studying cell-to-cell communication under a wide range of environmental conditions , since a cooperator under one condition might be a cheater under another . The population structure ( see figure 1 ) , which results in two levels of selection , was essential for the maintenance of the different ecotypes [66] , [73] , [74] . Previous studies have shown that population structure can facilitate the evolution and maintenance of cooperation [69] , [75]–[83] . The population structure makes individuals interact assortatively [84]: cooperators are therefore more likely to interact with other cooperators than cheaters . As a consequence , the benefits of cooperation mostly end up with cooperators , due to which there is a net selective advantage for cooperation . In our model the degree of assortment depends on the number of individuals that initialize a single colony ( ) or , in other words , on the strength of the recurrent population bottlenecks [85] , . We assumed that the colonies themselves are well-mixed , although within-colony structure—via the emergence of assortment—might have facilitated cooperation even more [87] , [88] . When signal production is too costly , cell-to-cell communication does not evolve , because the selective advantage of cheaters at the cellular-level cannot be compensated by the selective advantage of cooperators at the colony-level . It is important to notice that our model only included signal production costs , even though plausible arguments could be made that the maintenance costs of a communicative system should be considered as well [89] . However , we do not expect that including maintenance costs would affect our results , since both cheaters and cooperators need to have a communicative system—and hence carry the associated costs—to sense the quorum-sensing signal . Although our model is limited to sporulation , it could be extended to examine the role of cell-to-cell communication in the timing of other differentiation events as well , for example: motility , bioluminescence , conjugation , competence , matrix-production , biofilm formation , biofilm detachment , etc . ( e . g . [11]–[14] , [47] , [90]–[94] ) . Every time there is a trade-off between the growth rate of two cell types ( e . g . cells and spores ) over two or more environmental niches that alternate over time ( e . g . nutrient availability and nutrient scarcity ) , a cell has a selective advantage when it accurately times the developmental transitions between both cell types ( see also [45] ) . When the population size affects the optimal time at which a cell should differentiate ( e . g . when a cell must predict future nutrient conditions ) , cell-to-cell communication is expected to evolve in order to enhance a cell's developmental timing . The challenge for future studies is to unravel the developmental trade-off and ecological niches that underlie each of these differentiation events . Furthermore , our study emphasizes the importance of examining the integration of different environmental cues in cellular decision-making [49] , [95]–[97] . The quorum-sensing threshold—and hence the critical population density—at which a differentiation event occurs can and mostly will strongly depend on other environmental conditions , such as nutrient availability [34] , [48] , [98] , [99] . | Biological systems are characterized by communication; humans talk , insects produce pheromones and birds sing . Over the last decades it has been shown that even the simplest organisms on earth , the bacteria , communicate . Despite the prevalence of communication , it is often hard to explain how communicative systems evolve . In bacteria , communication results from the secretion of molecular signals that accumulate in the environment . Cells can assess the concentration of these signals , which indicate cell density , and respond in accordance . This form of cell-to-cell communication is responsible for the regulation of numerous bacterial behaviors , such as sporulation . Spores are metabolically inactive cells that are highly resistant against environmental stress . It is adaptive for a cell to sporulate when it struggles to survive . We show , via individual-based simulations , that cell-to-cell communication evolves because it allows cells to predict future environmental conditions . As a consequence , cells are capable of anticipating environmental stress by initiating sporulation before conditions are actually harmful . Furthermore , our model shows that cell-to-cell communication can even drive ecological diversification , since it facilitates the evolution of individuals that specialize on distinct ecological conditions . | [
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion"
] | [
"bacteriology",
"signaling",
"networks",
"evolutionary",
"selection",
"microbiology",
"cell",
"differentiation",
"developmental",
"biology",
"microbial",
"evolution",
"molecular",
"development",
"microbial",
"growth",
"and",
"development",
"theoretical",
"biology",
"evolutionary",
"modeling",
"biology",
"adaptation",
"signaling",
"computational",
"biology",
"evolutionary",
"biology",
"bacterial",
"evolution",
"evolutionary",
"processes"
] | 2012 | The Evolution of Cell-to-Cell Communication in a Sporulating Bacterium |
Convergent extension , the simultaneous extension and narrowing of tissues , is a crucial event in the formation of the main body axis during embryonic development . It involves processes on multiple scales: the sub-cellular , cellular and tissue level , which interact via explicit or intrinsic feedback mechanisms . Computational modelling studies play an important role in unravelling the multiscale feedbacks underlying convergent extension . Convergent extension usually operates in tissue which has been patterned or is currently being patterned into distinct domains of gene expression . How such tissue patterns are maintained during the large scale tissue movements of convergent extension has thus far not been investigated . Intriguingly , experimental data indicate that in certain cases these tissue patterns may drive convergent extension rather than requiring safeguarding against convergent extension . Here we use a 2D Cellular Potts Model ( CPM ) of a tissue prepatterned into segments , to show that convergent extension tends to disrupt this pre-existing segmental pattern . However , when cells preferentially adhere to cells of the same segment type , segment integrity is maintained without any reduction in tissue extension . Strikingly , we demonstrate that this segment-specific adhesion is by itself sufficient to drive convergent extension . Convergent extension is enhanced when we endow our in silico cells with persistence of motion , which in vivo would naturally follow from cytoskeletal dynamics . Finally , we extend our model to confirm the generality of our results . We demonstrate a similar effect of differential adhesion on convergent extension in tissues that can only extend in a single direction ( as often occurs due to the inertia of the head region of the embryo ) , and in tissues prepatterned into a sequence of domains resulting in two opposing adhesive gradients , rather than alternating segments .
Convergent extension refers to the simultaneous narrowing and extension of tissues due to extensive cell rearrangements , and is a key morphogenetic event during formation of the bilaterian body plan . In bilaterian animals , convergent extension first occurs when the main body axis forms and extends , pushing the head and tail further away from each other . Although this axis extension is universal in bilaterians , the cell and tissue behaviour causing it differs widely between species ( for reviews see [1–4] ) . In Xenopus for example , dorsal mesodermal cells polarize and change their adhesive properties ( reviewed by [5] ) ; cells then crawl between each other in a zipper-like process called intercalation [1 , 2] . In contrast , convergent extension of zebrafish mesoderm consists of two processes: directed migration to the dorsal axis and intercalation [2–3 , 6] . Finally , Drosophila germband extension occurs in a tightly connected epithelium , where cells intercalate by contracting those parts of the membrane that have a dorsal-ventral orientation [7–9] . Convergent extension is an inherently multiscale process , in which subcellular contractility and adhesion , cell level polarity and migration , and tissue level deformations are involved . Models incorporating this multiscale nature are of key importance to study the feedback interactions that give rise to tissue extension . Thus far , models are largely conceptual in nature , testing whether an experimentally observed ( sub ) cellular process or hypothetical mechanism can indeed drive convergent extension [10–13] . Among the identified mechanisms capable of driving convergent extension confirmed by these models are lamellipodia formation [13 , 14] , directed mitosis [13] , oriented membrane contraction [8 , 11] , cell extension or protrusions [10–12] and anisotropic differential adhesion [15] . These different mechanisms also differ in the origin of the directional signal , the cue which informs cells into which direction to move . In the models including either directed mitosis , lamellipodia or oriented membrane contractions [11 , 13 , 14] , this direction is explicitly imposed in the model by telling cells in which direction to extend . In contrast , in the models with anisotropic differential adhesion and cell elongation [12 , 15] , there is no global information: cells have internal polarity and through cell-cell interactions the cells align . Other models are somewhere in between; Rauzi et al . [8] use experimental data on the polar distribution of actomyosin , resulting in a coordinated contraction of only dorso-ventrally oriented membranes . The model by Weliky et al . [10] does not impose the direction in which cells extend , but includes two boundaries enclosing the tissue which inhibit cell extensions , thus providing an overall bias . Regardless of how cell/tissue polarity is incorporated in these models , convergent extension has so far always been studied in homogeneous tissues consisting of cells with identical fates . However , axis extension usually does not occur in homogeneous tissues , but rather in tissues that have been or progressively become patterned into regions of different cell fate . In Tribolium for instance , segments are formed by an oscillating gene clock , shortly after which the newly segmented part of the tissue starts to narrow and extend [16–19] . Therefore , an interesting question is how patterns are maintained under convergent extension , which leads to extensive cell rearrangements and therefore potentially mixes up cells of different fate . Considering this , it is striking that in Xenopus , the antero-posterior patterning of the mesoderm is crucial for convergent extension [20]; also in Drosophila , a segmented body pattern is essential for germband extension [21 , 22] . This leads to the intriguing suggestion that rather than segments becoming lost due to convergent extension , these segments may actively drive convergent extension . Since the interplay between tissue patterns and convergent extension has so far received little attention , we use a computational model to investigate how a segmented tissue pattern can be maintained during convergent extension , and whether and how such a pattern may itself drive convergent extension . We use the Cellular Potts model ( CPM ) formalism [23 , 24] , which has been successfully used to model different mechanisms of convergent extension in a homogeneous tissue [12 , 15] , as well as several other morphogenetic processes like somitogenesis [25] , ommatidia formation in Drosophila [26] , and Dictyostelium culmination into a fruiting body [27] . CPM is particularly suitable for performing the type of multiscale simulations necessary to investigate convergent extension since it endows cells with an explicit size and shape , allowing for both subcellular resolution and deformation , as well as cell level properties such as adhesion and migration [28] . In this work , we show that convergent extension by itself tends to disrupt a segmented gene expression pattern that was previously formed . We demonstrate that this disruption may be counteracted by letting cells adhere preferentially to cells of the same segment type . Furthermore , we find that such segment-specific adhesion by itself can both provide the directional signal and serve as a driving force for convergent extension . When we add a simple form of directional persistence ( representing inertia in the cell’s direction of movement due to the delay caused by cytoskeleton-recycling dynamics ) this substantially increases the efficacy of convergent extension through segment-specific adhesion . The latter is especially true in larger and stiffer tissues , where segment-specific adhesion alone is insufficient to cause a significant tissue shape change .
For all of our simulations we used a 2D CPM model with two different cell types ( red and green ) which represent segments with different identities . These in silico cell types can either have no segment-specific adhesion -a green cell will then adhere equally strongly to a red cell as to a green cell- or have segment-specific adhesion , meaning that cells prefer to stick to cells of the same type . In both cases , cells adhere more to other cells than to the surrounding medium , such that the tissue does not fall apart into separate cells or tissue types . The medium itself has no other properties than its adhesion with cells . In CPM , adhesion is regulated by J values , which represent the surface energy ( per amount of contact surface ) between cells of the same type , between cells of a different type , or a cell and medium . The CPM tries to minimize the total energy of the system , so contacts with lower J values are preferred . The strength of adhesion or repulsion between cells depends on the difference between J values , which can be conveniently represented by the surface tension ( γ ) [24] . The γ values are calculated from the J values as follows: γ i , j = J i , j − J i , i + J j , j 2 , where i and j represent different cell types ( m = medium , r = red , g = green ) . Note that Jm , m = 0 . We will refer to γ values throughout this paper , J values are mentioned in the figure legends . A positive γi , j value means that cells prefer to adhere to cells of the same type , whereas negative values indicate that cells of different types prefer to mix . We will only use positive or 0 values for γ . For a subset of simulations , we added a so-called persistence mechanism to our model . Persistence is the tendency of cells to maintain their previous direction of movement ( memory or inertia ) , due to the non-instantaneous turnover of the cytoskeleton [29] . When persistence is strong , cells are able to migrate rapidly in a consistent direction , as observed for example in lymphocytes and in gastrulating cells in zebrafish . We implemented persistence by giving cells a favoured ( target ) direction of movement . Despite this bias , the cell is not always able to move exactly in this direction due to hindrance by other cells or simply random fluctuations . Therefore , this target direction was regularly ( after a fixed number of simulation steps ) updated with the cell’s actual direction of displacement , representing the eventual remodelling of the cytoskeleton ( Fig . 1C ) [30] . A cell moving this way performs a persistent random walk . Initially , we also tested two explicit mechanisms for convergent extension , which both used global information to direct the cells . The first mechanism , called graded adhesion , was based on the observation that mesoderm cells in zebrafish follow a gradient in cadherin activity towards the central axis ( [31] , reviewed in [4] ) . In our model , we implemented this by imposing a static gradient of cell adhesion , where the location of the cell in the field determined how strongly it adhered to neighbouring cells; cell contacts that were closer to the center of the x-axis adhered more strongly than those that were farther away ( Fig . 1A ) . The second mechanism , called axial adhesion , was an adapted version of the mechanism presented by Zajac et al . [15] . This mechanism was based on the observation that intercalating cells in Xenopus are polarized and elongated , and the conjecture that these cells may also have a polarized distribution of adhesive molecules along their membrane . Fig . 1B shows the basic idea: the upper and lower sides of a cell ( defined by the y-axis of the field ) have a higher density of adhesion molecules than the left and right sides . A cell’s adhesion to a neighbouring cell is then a product of the local density of adhesion proteins on both cells , which is approximated by adjusting the J value ( we don’t explicitly model adhesion proteins ) . We chose the axes for adhesive density such that the tissue should extend perpendicularly to the segments . Later on , these two mechanisms were no longer applied . For a detailed description of the implementation of all mechanisms , we refer to the Methods section . We initiated the in silico tissues with a regular , segmented pattern of red and green cells ( Fig . 1D ) . For convenience , we use the terms anterior-posterior ( A-P ) axis and medio-lateral ( m-l ) axis when we talk about the major and minor body axes of the tissue ( which may have any orientation in the field ) . When we refer to the axes of the field ( which are fixed ) , we simply use x-axis and y-axis . In most simulations however , the y-axis and A-P axis had the same orientation , meaning that the tissue extended in the direction of the y-axis of the field . To study the effect of convergent extension on a pre-segmented tissue pattern , we started with the incorporation of either of the two explicit global mechanisms ( graded adhesion or axial adhesion ) without including segment-specific adhesion or persistence . We observed for both convergent extension mechanisms that the tissues extended and narrowed , but that cells at the boundaries invaded other segments , with some losing all contact with their designated segment ( graded adhesion , Fig . 2A , S1 Video; axial adhesion , Fig . 2D , S2 Video ) . Note that the strength of both mechanisms was relatively low in these tissues , and that the loss of segment integrity became more pronounced when the strength of the mechanisms was increased ( S1 Fig . ) . Next , we added a small positive surface tension between the red and green celltype ( γr , g = 4 , γc , m = 4 ) , causing preferential adhesion to same-segment-type cells . This sufficed to prevent cells from leaving their segment during convergent extension ( Fig . 2B and E; S3 , S4 Videos ) . Moreover , the boundaries of the segments were much straighter . To determine whether this differential adhesion caused any additional differences , we tracked the total direction of movement of each cell over the whole course of a simulation ( Fig . 2 , vector plots ) . In all cases we see the typical pattern of convergent extension: vectors directed inwards on the lateral sides , and outwards at the anterior and posterior ends . Despite the preservation of segments when segment-specific adhesion was present , there was very little difference in the appearance of the vectors . We colour-coded the displacement vectors according to the average angle with their neighbours ( Fig . 2A , B , D , E ) . It seems that in the presence of differential adhesion cell migratory dynamics are slightly more coherent . The vector plots in the cases with segment-specific adhesion ( Fig . 2B , E ) suggest that there was a considerable amount of A-P movement , which was unexpected given that cells remained restricted to their own segment . We checked whether this restriction led to a limitation of axis extension ( Fig . 2C , F ) . Strikingly , segment-specific adhesion did not limit axis extension , but in fact enhanced it . The fact that segment-specific adhesion seemed to enhance axis extension , ( Fig . 2C , F ) prompted us to investigate the effect of segment-specific adhesion without any additional mechanism for convergent extension ( Fig . 2G–I ) . Compared to tissue without segment-specific adhesion ( Fig . 2G ) , tissue that had a small amount of segment-specific adhesion ( the minimum amount needed to maintain segments in the presence of an explicit convergent extension mechanism ) , elongated significantly ( Fig . 2H ) . Furthermore , convergent extension occurred without the cells or tissue having an explicit notion of their A-P axis ( as opposed to the simulations in Fig . 2A–F , where the direction was imposed ) . This directionality now arose automatically , from the orientation of the interface between segments . The vector plot of the tissue with segment-specific adhesion ( Fig . 2H ) resembled the pattern generated by the graded and axial adhesion mechanisms , albeit with less extensive movement . This pattern was absent in the tissue without segment-specific adhesion ( Fig . 2G ) . These results showed that segment-specific adhesion may not only be able to maintain segments , but could also be a driving force of convergent extension on its own . To further investigate this possibility , we varied the surface tension between red and green segments ( difference in adhesion between like and unlike cells , γr , g ) , and the tension of cells with the medium ( γc , m ) . With increasing intersegment tension , having contact surface between segments becomes less energetically favourable , creating the tendency to reduce the segment interface . This caused the segments to round up more and so become narrower and thicker , so also the entire tissue extended and narrowed more strongly for increasing ( γr , g ) ( Fig . 3A ) . Moreover , the more the tissues extended , the more the vector plots in Fig . 3A resembled the typical pattern of convergent extension . Tissue extension is counteracted by increasing the tension with the medium , because extension and narrowing leads to a larger contact surface with the medium , which becomes less favourable with larger γc , m . Put differently: if cells prefer not to be in contact with the medium , the tissue as a whole will remain more rounded ( minimal surface with the medium ) and therefore extend less . The final amount of extension therefore depended on the balance between the two opposing tensions . For the case in the parameter space with the most extreme extension ( γr , g = 10 , γc , m = 6 ) , the tissue extended to about 1 . 5× its original length ( Fig . 3B , S5 Video ) . When we included more and thinner segments , the tissue extended even further ( to more than 2× the original length ) ; otherwise , the results were qualitatively similar ( box in Fig . 3 , S2 Fig . ) . Occasionally , we observed that two segments of the same celltype contacted each other and merged , thus reducing the number of segments ( Fig . 3A , bottom right; S2 Fig . , bottom row; S6 Video ) . This biologically unrealistic behaviour only occurred for very strong differential adhesion , while biologically relevant behaviour prevailed in the remaining , considerably larger part of the parameter space that we explored . So far , the in silico tissues with segment-specific adhesion reached their final length within about the same time scale as the explicit mechanisms . So far however , we used relatively small and loosely connected tissue . Therefore , we decided to investigate the efficacy of segment-specific adhesion in both larger and stiffer tissues . In larger tissues , cells would need to travel greater distances to achieve the same degree of extension; this could potentially mean that the same process takes much longer in a larger tissue . It has indeed been suggested that if surface tension alone had to drive large changes in tissue shape , the process would take unrealistically long [32] . In Fig . 4A we compare two in silico tissues with the same surface tensions and the same ratio between the length and width of a segment , but one consisted of four times more cells ( the number of cells in both the length and width of the segments was doubled ) . Because of the difference in total size , we used the ratio of the long axis over the short axis of the tissue to compare the extent of axis extension . It can be derived from first principles that for tissues with the same surface tensions and the same axis ratios at the start , the final axis ratio should be the same as well ( S1 Text ) . As expected , the larger tissue extended at a much slower pace than the small tissue and did not reach the same axis ratio within the span of the simulation . In the model we used so far , cells retained no memory of the direction in which they previously moved , and could change their direction of movement instantaneously . However , biological cells are to some extent persistent: due to polarization and turnover dynamics of the cytoskeleton they tend move for some time in a straight line before changing direction [29] . We hypothesized that endowing our in silico cells with some persistence in their movement might enhance the effectiveness of the cell motion resulting from segment-specific adhesion . Therefore , we implemented a simple persistence mechanism which has been used before in CPM for migrating lymphocytes [30] ( see section “the model” and Methods for details ) . Note that we did not impose a tissue-level bias on the direction of persistence beforehand to favour convergent extension: the cells started each with their own random target direction . Endowing cells with a limited tendency for persistence slightly increased the speed of cell displacement , yielding more rapid convergent extension and a more elongated tissue shape at the end of the simulation ( Fig . 4B , C ) . In the large tissue , further increasing the level of persistence allowed the tissue to reach almost the same axis ratio as the small tissue without persistence in a comparable amount of simulation steps ( Fig . 4C ) . The smaller tissue also gained extension speed and a larger axis ratio from increased cell speeds; however , because the tissue already extended quite rapidly , the contribution of persistence was substantially smaller ( Fig . 4B ) . From the vector plots it can be seen that the overall cell displacement pattern still generated the typical convergent extension pattern . Note that without differential adhesion between segments ( S3 Fig . ) , persistent cell motion only mixed up the segmentation pattern without yielding any tissue extension . This indicates that segment-specific adhesion provided the directional signal for axis extension; aligning the initially random direction of persistent cell motion and thus allowing it to enhance tissue extension . When the strength of the persistence mechanism was strongly increased , the probability of segments merging suddenly increased ( Fig . 4B , yellow curve ) . Interestingly , the large tissue seemed capable of sustaining larger cell speeds before segment collapse occurred . In both cases this extreme behaviour only occurred for rather strong persistence , while in a large part of the parameter space convergent extension was significantly enhanced without the risk of tissue collapse ( see also S4 Fig . ) . Next , we tested the efficacy of segment-specific adhesion in stiffer , more epithelium-like tissues . For this , we doubled the J values , which reduces the amount of membrane fluctuations; as can be seen in Fig . 4D and F , the cells move considerably less and have a more distinct hexagonal shape than in the more flexible mesenchyme-like tissue we studied earlier . In these stiff tissues , segment-specific adhesion alone generates hardly any tissue extension , because the higher J values present an energy barrier to tissue shape change , much like tight junctions ( Fig . 4D ) . However , combined with increasing levels of persistence , beyond the range of parameters used before , significant tissue extension arose ( Fig . 4E ) . Thus , in a stiff , tightly connected tissue an active cell motility process is required to drive convergent extension , while differential adhesion still provides the directional signal . Interestingly , for intermediate persistence levels cells maintain the hexagonal shape typical of stiffer epithelial tissues and T1 transitions can be frequently observed throughout the extension process ( Fig . 4F ) . In contrast , for the highest persistence levels tested , cells display considerable more membrane fluctuations , causing them to lose the hexagonal shape imposed by the higher tissue tension . Therefore , under these settings the tissue can no longer be considered as epithelium-like . The above simulations were all done with an unconstrained , fully segmented tissue . To further examine the relevance of differential adhesion as a driver of convergent extension in real-life morphogenesis , we modified our simulations in two ways corresponding to observed in vivo conditions: with a constrained anterior end and with a gradual instead of discretely segmented differential adhesion pattern ( Fig . 5 ) . In many cases , the tissue undergoing convergent extension is attached on one or more sides to adjacent tissue , and is therefore restricted in its movements in that direction . For instance , in Tribolium the converging tissue is attached to the head , which moves very little and does not change shape . We tested the influence of such a restriction on convergent extension and cell mixing by placing the anterior end of the tissue against the border of the field , to constrain tissue movement at the anterior tissue boundary . We then applied the explicit graded adhesion mechanism ( as in Fig . 2 ) , and observed that cell mixing still occurred in the absence of segment-specific adhesion ( Fig . 5A ) . Note that the anterior end of the tissue converged less because the tissue could not extend in the anterior direction , which becomes obvious in the vector plot of Fig . 5A where all arrows point either inward or to the posterior . This caused the tissue to become a bit ‘carrot-shaped’ , which is indeed typical for extending tissues attached to non-extending tissues ( see e . g . , Tribolium ) . Again , segment-specific adhesion prevented mixing at the segment boundaries ( Fig . 5B ) , and was by itself able to drive convergent extension ( Fig . 5C ) for larger surface tensions . Note that for strong segment-specific adhesion , the tissue tended to rotate and push away from the boundary to escape the restriction ( see also vector plot ) , allowing it to elongate more in the same amount of simulation steps . This is an artefact of the way we modelled the restriction as only an impenetrable boundary into which no extension can occur; had the extending tissue also been attached to this boundary it would likely rotate less . Convergent extension also occurs in non-segmented tissues . In Xenopus it was shown that when cells from the axial mesoderm were mixed , they quickly sorted out according to their original position on the antero-posterior axis , implicating that a position-based differentiation gradient rather than discrete segments yielded differential adhesion [20] . Strikingly , the amount of convergent extension occurring depended on the degree of sorting out that had already occurred . This suggested that differential adhesion , besides getting and keeping cells at the right position , also played an instructional role in convergent extension . Furthermore , it was found that the differential adhesion mechanism acted both upstream of and in parallel to the PCP pathway to drive convergent extension [20] . Here , we tested whether a gradient in adhesion proteins could cause correct anterior-posterior sorting and whether it could bring about convergent extension in a similar manner to the in silico segmented tissue . A tissue with graded expression of a single protein would not display anterior-posterior , but radial cell sorting without any convergent extension , according to both experiments and computational models [33 , 34] . We therefore generated a tissue with two adhesion proteins that formed opposite gradients . This meant that a cell with a high concentration of protein A had a low concentration of protein B and vice versa ( S5 Fig . , see Methods for details ) . Cells with high A adhere more strongly to other cells with high A ( and vice versa ) . Furthermore , cells with intermediate concentrations of both proteins adhere more strongly to each other than to cells with a high concentration of just one protein ( S5 Fig . , explanation in Methods ) . When cells were placed randomly in the tissue ( as in the experiment with mixed tissue ) , they sorted out with cells with similar protein concentrations clustering together . However , tissues in which cells had no persistence sorted out only partially: they became stuck in local optima where multiple clusters of similar protein concentrations were present , which was also observed for large tissues with a single protein gradient [34] ( S5 Fig . ) . The partially sorted state was reached more quickly when the gradients of A and B concentrations were steeper , although this still did not lead to complete sorting . The tissue did sort completely when cells were endowed with high persistence , creating a rather turbulent tissue which could sort quite rapidly , with high-A cells on one end and high-B cells on the other ( Fig . 6A ) . When the simulation started with a tissue in which cells were already sorted , it elongated , with the extent of elongation depending on the maximum difference in adhesion ( Fig . 6B , S6 Fig . ) . Modest persistence could enhance this process ( S6 Fig . ) , but strong persistence reduced the extension again ( see the fully sorted , but unelongated tissues in S5 Fig . ) . Therefore , if extension should follow after sorting of a fully mixed tissue , cell motility needs to be regulated together with the degree of sorting . However , in naturally occurring situations , AP patterning occurs prior to convergent extension , so complete mixing and hence the need for complete sorting are unlikely to occur . Rather , robustness to developmental noise will require limited sorting to optimize AP patterning , for which lower persistence levels are sufficient . Thus , our results show that besides a segmented tissue pattern , graded distributions of adhesion proteins are also capable of driving a modest form of convergent extension .
During formation of the bilaterian body axis , cells converge and intercalate to form a tissue that is longer and narrower . Convergent extension usually occurs in tissues which have undergone prior gene expression patterning such that cells have distinct fates at different positions in the tissue . Arguably , convergent extension , which often causes extensive cell rearrangements , should be tightly regulated to prevent it from interfering with this tissue pattern . An example where this is relevant is Tribolium , in which convergent extension follows shortly after segmentation [17] . Paradoxically , it has been shown in both Drosophila and Xenopus that a segmented or other antero-posterior tissue pattern is required for convergent extension [20–22] suggesting that it is instructive for rather than compromised by tissue remodelling . It is therefore important to know how convergent extension may interact with a prepatterned tissue . Here , we investigated the potential role of segment-specific adhesion in convergent extension of a fully segmented tissue . We applied two mechanisms -graded adhesion and axial adhesion- that caused convergent extension of the tissue . We demonstrated that without segment-specific adhesion , these mechanisms disturbed the segmented tissue pattern . Adding segment-specific adhesion in our model did not only preserve the segments , but also enhanced the extension of the long tissue axis . Furthermore , segment-specific adhesion by itself was sufficient for convergent extension both in unconstrained and constrained tissue , and can be combined with persistence to enhance extension in larger and stiffer tissue . Finally , we have shown that this differential-adhesion based mechanism also extends to non-segmented tissues with opposite gradients of adhesion proteins , although the amount of extension is more modest . An important question concerning convergent extension is where the directional signal for the orientation of tissue extension comes from . A number of earlier models was constructed to elucidate the various mechanisms behind convergent extension through cell intercalation in different organisms [8 , 10–13 , 15] . Most of these predefined a direction of extension either by biasing protrusions or constrictions of the cell membrane ( [8 , 11 , 13 , 14] ) , or including a boundary which restricts cell motion in certain directions [10] . Only two models did not impose such a direction . In the model by Backes et al . ( [12] , a positive tension between two cell types instructed the intercalation direction of forcibly elongated cells , and led to a direction of extension and narrowing which was perpendicular to that in our model . This mechanism only worked for tissues which were already quite narrow , and generated very little actual tissue extension . In the original version of the axial adhesion mechanism , constructed by Zajac et al . ( anisotropic differential adhesion , [15] ) , the adhesion polarity of cells was not fixed , but rather depended on the orientation of the cell long axis ( cells were forced to be elongated ) . In this case , the direction of tissue extension was not predefined , but emerged through alignment of the elongated cells . As a consequence , the direction of axis extension was random and differed between simulations . Finally , Shinbrot et al . [35] demonstrated that cell-cell adhesion and repulsion can generate segmented and elongated tissue patterns from random initial cell configurations . Rather than through convergent extension , the elongated and segmented patterns in their simulations form from cells condensing from a dispersed state while sorting into disks , with the tissue assuming a random orientation with respect to the field axes . In this paper , we started out with two superimposed mechanisms for convergent extension , in which the direction of extension was also superimposed . One had an explicit gradient defining the position of the extending axis ( graded adhesion ) , while the other imposed an internal , fixed polarity on the cells ( axial adhesion ) , thus implicitly assuming the presence of some kind of signalling gradient . Interestingly , when segment-specific adhesion drove convergent extension alone , directionality emerged without such an imposed signalling gradient or polarity . Instead , the interface between segments provided enough information to allow the tissue to stretch in the direction perpendicular to it . The ability of segment-specific adhesion to provide the extension direction was further emphasized when we combined it with the persistence mechanism , which by itself could not produce convergent extension , but could speed up tissue extension considerably when combined with segment-specific adhesion . Therefore , to our knowledge , segment-specific adhesion is the first convergent extension mechanism which yields a predictable direction of convergent extension without imposing polarity on the cellular or tissue level . In our model , the degree of tissue extension by segment-specific adhesion was determined by the balance between surface tension between red and green segments , and the tension of the tissue with the surrounding medium . The red-green surface tension provided the elongating force by reducing the contact surface between the segments ( pulling the segment interface inward ) , whereas the surface tension with the medium opposed this force by making the tissue as a whole as round as possible ( pulling the segment interface outward ) . This agrees well with findings in Xenopus , where the axial mesoderm needs to be enveloped in epithelium in order to extend [36] . Without the epithelial layer , the surface tension of the mesoderm with the environment is too high , resulting in a spherical tissue . Because differential adhesion minimizes the contact area between tissues of different types , the initial ratio between segment width and length is another factor influencing the extent of convergent extension in our simulations . The smaller the initial ratio between segment width and length , the larger the contraction of the contact between segments , and the more extreme the resulting tissue elongation will be ( compare Fig . 3 with ratio 3/10 , to S2 Fig . with ratio 2/15 ) . As segments are typically organized perpendicular to the length axis of the tissue , the initial segment width corresponds to the tissue width before convergent extension , while the initial segment length corresponds to the tissue length before convergent extension divided by the number of segments . The segment width length ratios used in our simulations are well within the naturally occurring ranges when considering the segment numbers and tissue widths and lengths observed in for example Tribolium , Drosophila and other arthropods . We observed an apparent limit to the extent in which segment-specific adhesion can drive convergent extension . When differential adhesion tensions or persistence levels exceeded a certain threshold , the segments started to rotate and merged with other segments of the same type , thus further minimizing intersegment boundary surface . However , we suggest that this may largely be an artefact of our simplified 2D model; the risk of tissue bending may be much lower for a 3D tissue , and/or if the tissue is also embedded in other tissues ( as in Xenopus ) that restrict its movements and aid convergent extension at the same time . Furthermore , the phenomenon did not occur for most of the parameter region we tested , and we obtained strong tissue elongation within the biologically relevant region . In addition , for persistence it is reasonable to expect that once convergent extension has completed cell motility is downregulated again as part of the further progression of the development program ( note that persistence is not required for maintenance of tissue extension ) . This termination of persistence provides an additional safeguard against segment fusion . In the current study we have shown that in a coherent , fully presegmented tissue , segment-specific differential adhesion is a suitable candidate mechanism both for maintaining segment integrity and driving convergent extension . We did not take into account other processes that may take place at the same time as convergent extension . For instance in Tribolium and other short-germ insects , the segments are laid down sequentially instead of simultaneously , from a growth zone where cell division provides a steady source of undifferentiated tissue . It appears that in this case , convergent extension occurs shortly after a new segment is laid down [16 , 17] . Based on preliminary results , we expect that segment-specific adhesion will also suffice to drive convergent extension during sequential segmentation , but given the complexity of the growth zone and segment-definition dynamics , it is beyond the scope of this article to investigate this . Furthermore , we assumed for the sake of simplicity that a cell’s adhesion is a fixed property . However , we recognize that this may not always be the case , for instance when cells change the concentration of adhesion molecules on their membrane in response to interactions with other cells that possess different ( concentrations of ) adhesion molecules ( see [33] ) . This may influence the ability of differential adhesion to drive convergent extension . We found that for loosely connected , mesenchyme-like tissues differential adhesion alone or combined with a limited persistence of motion can drive convergent extension . As such , we expect differential adhesion to contribute to axial extension in organisms such as Xenopus in which an antero-posterior pattern is present , and which indeed served as one of the inspirational starting points for this study . Possibly this mechanism also plays a role in short-germ insects such as Tribolium , which undergoes convergent extension simultaneously with segmentation , if the tissue emanating from the growth zone is indeed flexible enough . For stiffer tissues we found that in order to obtain substantial tissue extension , segment-specific adhesion needs to be combined with a significant level of persistent cell motion . Notably , the persistence alone would not produce any convergent extension , but requires differential adhesion to instruct and coordinate cell movement . Furthermore , the strength of persistence required for proper extension was so low that inspection by eye would most likely not reveal the presence of this mechanism in in vivo tissues , as the cell displacement is similar to that of tissues where cells are not persistent . Persistence strong enough to be visible led to turbulence and segment merging , and would require other , more global directional cues than segment-specific adhesion to yield convergent extension . Although clearly not a one-to-one match , persistence bears intriguing similarities to the case of the Drosophila germband . Here , parasegmental actomyosin barriers prevent intersegmental cell mixing [37 , 38] , while the segments also serve as a directional signal for planar cell polarity , which subsequently instructs the anisotropically directed actomyosin contractions that drive the T1 transitions underlying convergent extension [7–8 , 22] . The similarities reside in the fact that the segmental pattern instructs the direction of cell movement , and that cell movement requires active cytoskeletal remodeling . It remains to be established whether segment-specific adhesion can act in combination with and thereby enhance the mechanisms observed in Drosophila , or whether it may act as an alternative strategy deployed in other organisms . Unfortunately , the similarities between the differential adhesion mechanism and the Drosophila type mechanism make the design of an experimental setup to discriminate against these two possibilities highly non-trivial . As an example , if we experimentally disrupt genetic factors regulating segmentation these will not only hamper segment-specific adhesion , but also the aforementioned planar polarity such as occurs in Drosophila . As a consequence results would be inconclusive . Likewise , active cytoskeletal dynamics are involved both in convergent extension driven by the combination of differential adhesion and persistence and in planar-polarized junctional tension driven convergent extension . Thus , failure of convergent extension upon actomyosin disturbance will again be inconclusive . Similarly , although one could try to experimentally increase the adhesiveness of the whole tissue by ubiquitously expressing e . g . N-cadherin; this would certainly hinder convergent extension via segment-specific adhesion , but unfortunately is likely to also hinder other convergent extension mechanisms by increasing the energy required to break the bonds between cells . This problem of distinguishing between the two mechanisms is further aggravated by the fact that the mechanisms may be likely to work in combination . One experiment that may allow for a distinction between the two convergent extension mechanisms is to apply pulling forces on the tissue in the direction parallel and perpendicular to the segmentation pattern . If less force is required to tear the tissue along segment boundaries than to tear it in the perpendicular direction this is a strong indicator that differential adhesion is involved . Still , this does not allow one to establish the importance of this differential adhesion for convergent extension . In summary , we have shown that differential adhesion is sufficient to drive convergent extension in presegmented tissues , and represents a convergent extension mechanism not requiring any directional signal . While the investigated convergent extension mechanism may not be universal , in segmented tissues the presence of segmental boundaries is likely to contribute to convergent extension , either via differential adhesion or via alternative mechanisms such as actomyosin bands or planar cell polarity . Likewise , while not all tissue is segmented , anterior posterior patterning may also allow for differential adhesion-based convergent extension . In the current study we focused on the role of a fully presegmented tissue pattern in driving convergent extension . However , in many cases segmentation and convergent extension occur simultaneously . Therefore , in future work we aim to investigate the dynamic interplay between sequential segmentation and convergent extension . Considering such bidirectional feedback between patterning and morphogenesis may bring to light important principles of coordinating growth and patterning .
Two alternative mechanisms for convergent extension were used: graded adhesion and axial adhesion . To implement such a mechanism , ΔH is modified for certain lattice sites to establish a bias in copy probability . The current study serves as a proof of principle , illustrating how convergent extension may disrupt pre-existing tissue patterns , while these pre-existing tissue patterns may also drive convergent extension through differential adhesion . Because of the conceptual nature of our model , we do not aim to quantitatively fit convergent extension dynamics in a particular model organism . However , if this where to be the case , model parameters could be adjusted to obtain cell movement speeds and trajectories matching experimental data . In contrast , in the current study we aim to illustrate that differential adhesion either alone or combined with persistence of motion , represents a feasible new mechanism for convergent extension . As such , we aimed to ensure that differential adhesion driven convergent extension occurs for a wide range of parameters , making it a plausible mechanism in broad range of contexts . For persistence , parameter scaling was done internally: we matched persistence tendencies to membrane fluctuations and overall tissue deformation so that we remained in the domain of biologically realistic behaviour , avoiding the merging of segments or of turbulent tissue dynamics . As a consequence , we applied considerably lower persistence tendencies than in the study by Beltman et al [30] , where it was used to simulate migrating lymphocyte dynamics . Note the limited persistence tendencies applied in our studies significantly altered quantitative model behaviour , as shown in Fig . 4 . Default parameter values are shown in Table 1 . | The process of convergent extension is a major contributor to the formation of the anterior-posterior body axis in the early embryo . Convergent extension refers to the directed movement of cells that leads to the extension of tissue in one direction and narrowing of the tissue in the perpendicular direction . Often , convergent extension occurs in tissue which already contains distinct domains of gene expression such as segments , and it is unclear how these patterns are maintained despite extensive cell movement . Interestingly , experimental evidence suggests that these tissue patterns may drive rather than be compromised by convergent extension . However , existing computational models aimed at unravelling the mechanisms of convergent extension have thus far only studied the process in homogeneous tissues . With our model , we demonstrate that in a segmented tissue , preferential adhesion of cells to other cells within the same segment type is required to maintain the tissue pattern during convergent extension . Furthermore , such segment-specific adhesion is by itself capable of driving convergent extension . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Segment-Specific Adhesion as a Driver of Convergent Extension |
In colonies of the honeybee Apis mellifera , the queen is usually the only reproductive female , which produces new females ( queens and workers ) by laying fertilized eggs . However , in one subspecies of A . mellifera , known as the Cape bee ( A . m . capensis ) , worker bees reproduce asexually by thelytoky , an abnormal form of meiosis where two daughter nucleii fuse to form single diploid eggs , which develop into females without being fertilized . The Cape bee also exhibits a suite of phenotypes that facilitate social parasitism whereby workers lay such eggs in foreign colonies so their offspring can exploit their resources . The genetic basis of this switch to social parasitism in the Cape bee is unknown . To address this , we compared genome variation in a sample of Cape bees with other African populations . We find genetic divergence between these populations to be very low on average but identify several regions of the genome with extreme differentiation . The regions are strongly enriched for signals of selection in Cape bees , indicating that increased levels of positive selection have produced the unique set of derived phenotypic traits in this subspecies . Genetic variation within these regions allows unambiguous genetic identification of Cape bees and likely underlies the genetic basis of social parasitism . The candidate loci include genes involved in ecdysteroid signaling and juvenile hormone and dopamine biosynthesis , which may regulate worker ovary activation and others whose products localize at the centrosome and are implicated in chromosomal segregation during meiosis . Functional analysis of these loci will yield insights into the processes of reproduction and chemical signaling in both parasitic and non-parasitic populations and advance understanding of the process of normal and atypical meiosis .
In most colonies of the honeybee Apis mellifera , the queen is the only reproductively active female and it produces pheromones that inhibit ovary activation in the workers [1 , 2] , keeping them sterile . Under some circumstances , particularly when the colony lacks a queen , workers may activate their ovaries and lay unfertilized haploid eggs via a process called arrhenotokous parthenogenesis . Due to the haplodiploid sex determination of honeybees ( and other hymenoptera ) these haploid eggs develop into male drones . The Cape bee ( Apis mellifera capensis Escholtz ) is a subspecies of the honeybee Apis mellifera that inhabits the far south of South Africa [3–5] . The Cape bee differs from other honeybee subspecies in that unmated females , particularly workers , frequently produce female progeny that develop from unfertilized eggs through thelytokous parthenogenesis . Thelytoky is a form of meiosis whereby two daughter nuclei that form in the oocyte fuse to produce a viable diploid egg [6] . This abnormal mechanism of restoring diploidy in unfertilized eggs occurs only occasionally in normally arrhenotokous populations ( <1% of eggs ) but is the norm in reproducing Cape bee workers ( >99% of eggs ) [7] . The Cape bee exhibits a number of other traits that distinguish it from other subspecies and lead it to behave as a social parasite , whereby workers invade the nests of foreign colonies , reproduce and utilize their resources . Cape bee workers have significantly larger ovaries that are more readily activated and contain more egg-producing ovarioles than worker bees of other subspecies ( 10–20 vs . 1–5 ) [3 , 8] . Parasitic egg-laying Cape bee workers produce queen pheromones , allowing them to assert reproductive dominance over other workers [9] and have a significantly increased lifespan of 3–5 months compared to 6 weeks in non-parasitic workers [10] . These traits likely increase the evolutionary advantage of parthenogenesis , which allows this mode of selfish reproduction to become established . As with other honeybee subspecies , Cape bee workers rarely reproduce if a queen is present ( ~0 . 1% of drones are produced from worker-laid eggs in a normal honeybee colony [11] ) . However , during reproductive swarming in Cape bees as many as 10% of young workers are produced by egg-laying workers [12] . In between 40% to 60% of young queens produced during swarming are the offspring of workers [13 , 14] . The genetic basis of thelytoky and its associated behavioral and physiological traits in the Cape bee is still largely unclear . Asexual reproduction by thelytokous parthenogenesis is relatively common in other species of Hymenoptera [15] . In the parasitoid wasp Lysiphlebus fabarum , where different forms exhibit arrhenotokous or thelytokous parthenogenesis , a single locus determines this difference [16] . Thelytokously reproducing individuals of this species are all homozygous for an allele at a particular microsatellite locus , which is very rare in arrhenotokously reproducing individuals . In the Cape bee , the genetic basis of thelytoky has been investigated using backcrosses of hybrids formed by crossing A . m . capensis with A . m . carnica [17] . The pattern of segregation in such crosses is generally consistent with inheritance of thelytoky being determined by a single recessive locus , which has been termed thelytoky ( th ) , although more complex modes of inheritance incorporating more loci are also compatible with this inheritance pattern . This putative locus has been mapped to an interval on chromosome 13 [18] and a 9 bp deletion in this region has been proposed as the causative variant [19] , however , a subsequent study failed to replicate this association [20] and detected the presence of the deletion in populations of other subspecies where thelytoky is absent ( see Materials and Methods for more details ) . Cape bees are genetically extremely similar to other African subspecies of honeybees , including A . m . adansonii and A . m . scutellata [21] suggesting a recent common ancestor and/or high levels of gene flow . Social parasitism in Cape bees should therefore represent a recently derived adaptation that has likely experienced positive selection . The distribution of Cape bees shows strong concordance to the biodiverse Fynbos ecoregion of South Africa , which suggests that the traits specific to the Cape bee are evolutionarily advantageous in this region [5] . Cape bee workers reproduce by thelytokous parthenogensis in the absence of a queen , during reproductive swarming or when invading the colony of a different subspecies . The propensity to do this is a trait that is fixed or close to fixation in the Cape bee population [3 , 22] . However , in their native range most reproduction in Cape bees is sexual , so patterns of genetic diversity therefore resemble those in other sexual outcrossing populations [21] . We therefore expect genetic variants responsible for traits associated with social parasitism to be located in regions of high differentiation with populations of other subspecies and associated with signals of selection in Cape bees . What is the genetic basis of social parasitism in Cape bees ? Social parasitism is facilitated by a range of different traits related to ovary development , behaviour and abnormal meiosis . Does a single thelytoky locus act as a “master switch” to enable this suite of phenotypes as previously suggested [17–19] , or are each of these traits determined by an independent set of loci ? To what extent are Cape bees genetically distinct from other African honeybee populations ? Was the emergence of social parasitism associated with increased positive selection in the Cape bee population ? Here we use population-scale genome sequencing to address these questions . We developed a method to detect selection in Cape bees compared to other African populations that combines detecting SNPs with high differentiation between populations with long haplotype tests of selection [23 , 24] . Here we apply this approach to a dataset of 10 whole genome sequences of Cape bees , sampled from locations were workers are known to produce nearly exclusively female offspring [25] , compared to 20 genome sequences of other sub-Saharan African honeybees that do not reproduce thelytokously . We detect a number of genomic regions with strong signals of selection in the Cape bee , which are likely to underlie the unique suite of traits that facilitate social parasitism in this subspecies .
We aimed to identify genetic variants associated with strong differentiation between the Cape bee and other African bees due to selection in the Cape bee in favor of traits involved in social parasitism . We analyzed whole genome sequences from 10 Cape bees ( A . m . capensis ) sampled from two localities in southern South Africa: Port Elizabeth to the east ( n = 5 ) and Cape Town to the west ( n = 5 ) . We compared these with sequences from two other African subspecies: A . m . scutellata collected in Pretoria ( n = 10 ) and A . m . adansonii collected in Nigeria ( n = 10; Fig 1 ) . Genetic differentiation between these three African populations ranges between FST = 0 . 051–0 . 056 ( Fig 1 ) and A . m . scutellata appears to be nearly equidistant genetically between the Cape bee in the south and A . m . adansonii in the north . This level of divergence is consistent with a recent common ancestor and/or pervasive gene flow among sub-Saharan populations . We considered the geographically widespread A . m . scutellata and A . m . adansonii populations as a single African background population and attempted to identify alleles under selection in the Cape bees compared to this background population . FST between the Cape bee population and the background population is 0 . 044 and the allele frequency spectrum is dominated by variants segregating at similar frequencies: 93 . 3% of SNPs ( n = 5 . 89 x 106 ) have FST values below 0 . 1 , whereas only 0 . 33% of SNPs ( n = 20 , 460 ) have FST values above 0 . 3 ( S1A Fig ) . There are 45 SNPs fixed for different variants between the Cape bees ( n = 10 ) and other African bees ( n = 20 ) , which are associated with one gene accession on chromosome 1 corresponding to Ethr ( ecdysis-triggering hormone receptor ) and three predicted genes on chromosome 11 ( hypothetical proteins GB44917 , GB45238 and GB45239; S1 Table ) . When the two Cape bee subpopulations are compared against other African populations separately , we observed 423 such fixed SNPs in the ( western ) Cape Town population ( n = 5 ) , which are associated with 38 gene accessions ( S1B Fig; S1 Table ) . However , in the ( eastern ) Port Elizabeth population ( n = 5 ) , we detect only 60 fixed SNPs ( associated with five gene accessions ) . FST between the African background population and the Cape Town population ( 0 . 070 ) is slightly higher than the African background population and the Port Elizabeth population ( 0 . 063 ) . The Cape Town population also has a general excess of high FST variants . Between the two comparisons against the African background population , the analysis with the Cape Town subpopulation include 70% or more of the variants segregating at FST>0 . 5 ( S1B and S1C Fig ) . Colonies with the entire suite of Cape-bee-specific phenotypes are more common in the Cape Town region , whereas they occur less frequently in Port Elizabeth [3 , 22] and hence the higher number of fixed SNPs in the Cape Town region may reflect additional variants associated with these phenotypes . Natural selection is expected to drive adaptive genetic variants towards fixation together with linked neutral variation , generating a pattern of increased population differentiation and higher haplotype homozygosity around selected variants [26] . We sought to identify genetic variants with high genetic differentiation due to recent selection specifically in the Cape bee population , reasoning that such variants likely underlie traits associated with social parasitism . To do this , we integrated SNP-level FST estimates with haplotype structure analyses . We implemented the population branch statistic ( PBS ) [27] to detect genomic regions that diverged rapidly on the Cape bee lineage compared to other African populations . We also scanned genome variation for genetic variants linked to long haplotypes in the Cape bee population using the cross-population extended haplotype homozygosity statistic ( XP-EHH ) [23] ( implemented in the program selscan [28] ) . For both PBS and XP-EHH estimates , values greater than zero are associated with selection signals in the Cape bees , whereas negative values indicate selection in the African background population . Across all SNPs in the dataset we find that the distribution of these values are centered around zero ( mean XP-EHH = 0 . 007; mean PBS = 0 . 039 ) , but that the most extreme XP-EHH and PBS scores are biased towards stronger signals in the Cape bees rather than the background population: for the PBS , the upper and lower 99 . 9% percentiles of the empirical distributions are 1 . 27 and -0 . 34 , respectively , and for the XP-EHH theses values are 4 . 12 and -3 . 46 ( S2A and S2B Fig ) . Low FST SNPs ( FST<0 . 3; >99% of SNPs ) appear to be largely unbiased between the two groups with respect to the PBS and XP-EHH statistics ( scores are close to zero ) , whereas high FST SNPs tend to have high values of these statistics in Cape bees: the average PBS and EHH scores of SNPs with FST>0 . 85 are within the top 0 . 5% of these statistics ( p<0 . 01 , bootstrap; S2C–S2E Fig ) . Moreover , variants segregating at the highest frequency differences are not randomly distributed but tend to be clustered in regions with the strongest linked signatures of selection: regional FST drops to the top 1% level after about 50kbp for SNPs at FST>0 . 9 , compared to about 20kbp for SNPs at FST = 0 . 8–0 . 9 , and these patterns decay similarly for PBS and XP-EHH ( S3A–S3C Fig ) . These analyses suggest that more variants are associated with selection signals in the Cape bee population compared to the background population . Many of these variants may underlie the unique set of derived characteristics associated with thelytokous parthenogenesis and social parasitism in the Cape bee . In order to test the robustness of these observations , we performed two additional reciprocal comparisons: one comparing A . m . scutellata to the remaining two African populations , and the other comparing A . m . adansonii to the remaining populations . For each of the comparisons , we estimated FST , PBS , and XP-EHH statistics in the same way as in the original genome scan ( Fig 2A ) . The distribution of FST in the additional comparisons has fewer peaks than in the Cape Bee , indicating fewer signals of selection specifically in these populations ( SNPs with FST>0 . 8 occur on 13 chromosomes comparing Cape bees against the other African populations , but only on one and six chromosomes respectively in the corresponding A . m . scutellata and A . m . adansonii scans; S4A–S4C Fig ) . In some cases we detected partially overlapping signals ( for instance around position 14 , 000 , 000 on chromosome 1 ) such that some accessions had outlier SNPs associated with more than one subspecies . However , among the gene candidates identified ( see section Characterization of candidate genes below ) , 95% and 99% of the accessions had higher overall PBS signals in the Cape bees than in A . m . scutellata or A . m . adansonii , respectively , consistent with an excess of uniquely derived variants and stronger selection in the Cape bees in most cases . In the comparison with A . m . scutellata , there are extremely few SNPs with high FST ( FST>0 . 80; 7 SNPs , compared to 730 SNPs in the Cape bee comparison ) . These SNPs have mildly elevated PBS values , but do not appear to be associated with signals of selection in A . m . scutellata according to the XP-EHH test ( Fig 2B–2D ) . In the comparison with A . m . adansonii , a large number of SNPs have high FST , but they are restricted to a few genomic regions . It is likely that many of these SNPs lie within regions of structural variation such as inversions that differ in frequency in the A . m . adansonii population , as high-FST SNPs are restricted to a few large blocks in this comparison . For instance , 84% of 1 , 241 SNPs with FST>0 . 8 are located to two 0 . 5Mb blocks on chromosome 7 and a 1Mb block on chromosome 11 ( S4C Fig ) . However , PBS is only marginally biased towards increased divergence in A . m . adansonii at high-FST SNPs in this comparison . Furthermore , XP-EHH values are below zero , indicating that differentiation at these variants is more strongly associated with selection in the background population ( A . m . capensis and A . m . scutellata ) ( Fig 2B–2D ) . Taken together , these observations confirm that despite similar levels of genetic differentiation overall , the number of genetic variants that are highly differentiated due to selection in the Cape bee population is much higher than in other African populations . These genetic variants are therefore likely to underlie the suite of traits connected to social parasitism . Since the western Cape Town and eastern Port Elizabeth Cape bee subpopulations showed different levels of fixation against the background population , we also queried these two populations separately for linked signals of haplotype divergence and diversity using three sets of outlier SNPs: those identified as 0 . 1% outliers for FST against the background population and with minimal evidence for extended haplotype homozygosity in the Cape bees ( XP-EHH>0 ) across the complete dataset and those identified for the same criteria , but for each subpopulation individually . The associated haplotype-level signals are consistently stronger in the western Cape Town subpopulation compared with the Port Elizabeth subpopulation for all three sets ( S5A–S5C Fig ) , congruent with greater fixation of Cape-bee-specific traits in the western region [3] . In order to identify variants specifically under selection in Cape bees , we estimated a composite selection score ( CSS ) [24] based on FST and XP-EHH scores ( see Materials and Methods ) . We did not include PBS in this score as the magnitude of PBS is expected to correlate with FST under neutrality . However , a strong association between FST and XP-EHH is not expected according to the results of coalescent simulations [29] . In our dataset , we find a positive and significant but weak correlation between the absolute values of the XP-EHH statistic and FST ( R2 = 0 . 07; p<1e-5 ) , consistent with the simulated data . The genome-wide distribution of FST , XP-EHH and PBS are shown in Fig 3A–3C . The distribution of CSS scores of all six million SNPs is shown in Fig 3D . We found that SNPs with the top 2000 CSS scores were significantly enriched within coding sequence by a factor of 2 . 3x ( p<1e-5; Fischer’s exact test ) compared to the expected . Likewise , SNPs are enriched within UTR regions and introns by 1 . 9x and 1 . 3x respectively ( p<1e-5 in both cases ) . Taken together , this enrichment indicates that extreme signals are centered on gene bodies and that coding variants in particular are more commonly targets of selection than non-coding intergenic ones , which are underrepresented in the top SNPs ( 0 . 57x the expected amount; p<1e-5 ) . Due to the association of outlier SNPs with gene bodies and the difficulty in characterizing the functional relevance of intergenic SNPs in gene deserts , we focused downstream analyses on outlier variants located close to genes . We selected a conservative set of top-ranking SNPs to produce a set of putative candidate genes . Among the top 1000 SNPs in the full dataset , 873 SNPs ( 87 . 3% ) occur within 8kbp of a gene , compared to 72% overall , a gene-centric overrepresentation ( p<1e-5 ) consistent with the enrichment reported above . We selected these 873 SNPs plus the next 137 for analyses based on the top 1000 SNPs located within 8 kb of a gene ( CSS = 3 . 04–5 . 52; 0 . 022% of all SNPs <8 kb from a gene; S2 Table ) . These make up 87 . 7% of the top 1140 SNPs . The average FST of these SNPs is 0 . 67 , indicating that although they were selected on the basis of both FST and XP-EHH , they typically have high allele frequency differences between the Cape bees and the African background population . Indeed , among these SNPs , the average frequency of the major allele in the Cape bees is about 90% , compared to only 20% in the African background and 18% in the European sample used as outgroup for the PBS ( S2 Table ) . The frequency of the major Cape bee allele was below 0 . 2 in 63% and 80% of SNPs in the African background and European populations , respectively . This demonstrates that putative candidate variants for social parasitism are typically rare outside of the Cape bee population . We next queried the 1000 SNPs against the current honeybee gene annotation models and found them to be located to peaks that associate with 97 gene accessions ( S3 Table ) . These gene accessions have average CSS levels across the full gene bodies within the top 1% percentile of genes ( CSS97 = 1 . 654 vs CSS1% = 1 . 546; S6 Fig ) . We queried the Drosophila orthologues ( n = 61 ) of this gene set for enriched gene ontology terms using the GOrilla platform [30] and detected 16 significantly enriched biological processes and molecular functions after controlling for multiple testing [31] ( p<0 . 05; S7A and S7B Fig ) . We used the REVIGO tool [32] to reduce redundancy in the list of GO terms , removing 4 terms ( S4 Table ) . Among the remaining GO-terms , we detect significant enrichment of genes associated with ecdysteroid metabolic processes ( GO:0042445; n = 5; p = 6 . 15e-3; S4 Table ) . The genes we detected within this category include orthologs of three glucose-methanol-choline ( GMC ) oxidoreductases ( GMCOX4/6/9 ) located in a conserved cluster on chromosome 1 [33] , and also Npc1a and foxo ( S4 Table; S7A Fig ) . Ecdysteroid signaling is a hormonal process that is correlated with development and ovary maturation in honeybees [34–36] . The corresponding top 1000 SNPs from the two other reciprocal comparisons ( scutellata vs the rest and adansonii vs the rest ) did not reveal any significantly enriched biological process GO terms . It is possible that genes associated with particular biological functions may be under selection but do not have clearly identifiable outlier variants under our criteria . In order to address this question , we analyzed a ranked list of all accessions with Drosophila orthologues ( n = 6 , 253 ) . We used the gene-wide average CSS scores to rank all genes and provided the list to GOrilla to perform analyses of enriched GO terms among top accessions while controlling for multiple testing . We did not detect any significantly enriched GO term with FDR-corrected p<0 . 05 using this method . To further address whether selection may be targeting genes specifically associated with meiosis and abnormal chromosome segregation ( i . e . thelytoky candidates ) , we collected all genes associated with GO terms related to meiosis or chromosome segregation . Genes of GO terms matching key words “meiotic” or “meiosis” ( 56 GO-terms; 126 genes ) and “segregation” ( 10 GO-terms , 74 genes ) were compiled into a meiosis-list . Each gene was counted only once despite the number of terms it was associated with , resulting in a list of 161 genes . In the same way , we compiled a list of 55 ecdysteroid-related genes ( i . e . ovary development candidates ) , associated with GO terms matching “ecdysteroid” ( 10 GO-terms; 20 genes ) or “ecdysone” ( 12 GO-terms , 41 genes; ecdysone is an ecdysteroid hormone ) . There is only a single shared gene between the two lists , so effects of multiple testing should be expected to be negligible . The average CSS score among all honeybee accessions is 0 . 383 . The mean CSS of the “meiosis-genes” is 0 . 377 . After randomly sampling 161 genes across the genome 2000 times we find that this minor decrease of about 1 . 5% is not significant at the 0 . 05 level . For the “ecdysteroid-genes” the mean CSS is 0 . 410 , a small increase of about 7% but not significant at the 0 . 05 level using the same resampling procedure . These alternative GO analyses fail to detect broad patterns of selection on particular biological functions . For instance , while we find that some of the clearest signals of selection based on outlier variants are associated with a small subset of genes involved in ecdysteroid metabolism , we do not find evidence for widespread selection signals across all genes associated with ecdysteroid metabolism . Taken together , the results of the GO analyses are consistent with a scenario in which selection in the Cape bees has targeted a restricted set of loci ( i . e . <1% of genes ) with diverse functions , rather than broadly affecting genes with similar functions across the whole genome . We further filtered the set of 97 candidate genes down to the subset of 25 accessions that had at least one top CSS SNP with an FST>0 . 8 ( nSNPs = 393; Table 1 ) . At these SNPs , the average frequency of the major allele in the Cape bees was 92% , but only 1 . 4% or 1 . 9% in the African background or European samples , respectively ( S2 Table ) . The frequency of the major Cape bee allele was found to be below 0 . 2 at >97% of these SNPs in either of the two populations . These 25 genes therefore contain the most highly differentiated variants with signals of selection between the Cape bees and the other African bees and those variants are also typically very rare also outside of Africa . Out of the original top 1000 SNPs , 762 SNPs are associated with these 25 genes and the average CSS score across these genes are within the top 0 . 5% of all genes in the genome ( CSS25 = 2 . 175 vs CSS0 . 5% = 1 . 912; S6 Fig ) , indicating that these genes have comprehensive signals of selection . Of these 25 accessions , 16 genes belong to the set of 38 accessions with fixed variants between the western Cape Town subpopulation and African background population and 4 genes belong to the set of 5 with fixed variants between the eastern Port Elizabeth and the background population . These are listed in Table 1 and highlighted in blue in Fig 3A–3D ( see S8 Fig for candidate gene plots ) . The top 762 SNPs associated with the 25 genes include 21 non-synonymous variants , affecting the coding sequence of 11 genes . Some of the 25 genes are located close together such that selection signals appear to extend across more than one accession . From manually inspecting these data we assigned these 25 genes into 12 distinct putative selective sweeps with strong evidence of selection in the Cape bee , including highly differentiated SNPs , which we labeled as A–L according to their location in the genome ( Table 2 ) . Corresponding plots of XP-EHH , FST , PBS , and the combined CSS around these sweeps are shown in S8 Fig . The three putative selective sweeps with highest CSS values are on chromosomes 11 ( locus H ) , 1 ( locus B ) , and 6 ( locus D ) and overlap with 14 of the accessions , with outlier signals spanning at least three genes each . The three SNPs with the highest CSS scores in the dataset are located in locus H on chromosome 11 ( ~40kbp wide ) , which contains 36 of the 45 fixed SNPs in the Cape bee sample relative to the African background population . The CSS scores are centered on the gene accession GB45239 , which contains the non-synonymous SNP with the highest selection score in the dataset ( CSS = 4 . 07 ) . This gene is not characterized in the honeybees but has peptide sequence similarity to a structural maintenance of chromosomes ( SMC ) domain ( COG1196; p = 6 . 82e-04; TIGR02168; p = 2 . 43e-03 ) in the NCBI Conserved Domain Database [37] . The role in chromosome integrity may indicate that this gene is involved in thelytokous meiosis in the Cape bee . Locus B on chromosome 1 is ~110kbp wide and encompasses 7 genes . However , the 9 fixed SNPs are centered at an intron of the Ethr gene , which codes for the receptor for the neuropeptide ecdysis-triggering hormone ( ETH ) . ETH is essential for initiating ecdysis but also regulates juvenile hormone ( JH ) synthesis in insects [38 , 39] . Notably , the region also spans genes coding for a StAR-like protein , an intracellular cholesterol transporter facilitating ovarian steroid hormone synthesis in vertebrates and insects [40–44] , and ebony , which makes N-β-alanyl-dopamine from dopamine , a process associated with yellow cuticular pigmentation instead of the brown/black tan resulting from dopamine melanization [45 , 46] . This region therefore contains candidates for larval development , ovary development and pigmentation , all of which differ in the Cape bee compared to other subspecies of A . mellifera . Locus D on chromosome 6 ( ~40kbp wide; 5 genes ) is centered on β-glucosidase , a digestive enzyme associated with carbohydrate metabolism , including cellulose digestion . In the honeybee , the gene codes for an enzyme that hydrolyses small mono-glycosides , with a particular preference for phenyl-glycosides such as β-pNPG [47–49] . While the EHH signal decays rapidly around the β-glucosidase gene , neighboring SNPs around the accession also have unusually high FST values . To either side of the accession are the UDP glycosyltransferase ( UGT ) genes encoding the two proteins UDP-glucuronosyltransferase 1-3-like and UDP-glucuronosyltransferase 2B13-like . UGT enzymes catalyze reactions that metabolize oligosaccharides and conjugate sugars to a broad range of lipophilic compounds ( see GO analysis results above ) . β-glucosidase has been shown to have a role in chemical signaling in social insects [50–52] . In the noctuid moth Mamestra configurata , β-glucosidase produces a glucoside precursor of the phenylethanol sex pheromone [50] . 2-phenylethanol has been suggested to advertise ovary status between young unmated honeybee queens and workers during the queen elimination phase [53] , but the role of β-glucosidase for producing the pheromone has not yet been studied in the honeybee . We find additional signatures of selection on loci outside of these three largest sweeps . High CSS scores across accession GB40769 “dehydrogenase/reductase SDR family member 11-like” ( LOC412458 ) on chromosome 1 ( locus A ) . SDRs make up large families of NADPH-dependent oxidoreductases with central metabolic functions [54] . The Drosophila orthologue is CG9360 , a NADP+-dependent farnesol dehydrogenase ( FOHSDR; KEGG: K15890 ) [55] . The orthologue detected in the closely related honeybee Apis florea using BLASTp [56 , 57] is also annotated as “farnesol dehydrogenase-like” ( E-value = 1e-140; 88% identity; XP_003690738 . 1 ) . This gene may therefore code for a farnesol dehydrogenase enzyme in bees . FOHSDR acts in the JH biosynthesis pathway and oxidizes the sesquiterpene alcohol farnesol into farnesal , a precursor of JH [39 , 58] . JH has an important function regulating ovary development and transition to adulthood [36] . Locus C , on chromosome 2 encompasses GB50742 ( LOC102655146 ) , annotated as “Bardet-Biedl syndrome 1 protein-like” ( BBS1 ) , a member of the BBSome protein complex , which is an essential protein transporter for cilium assembly and localizes near the primary cilium and centrosome in the cell [59 , 60] , which suggests it may be a candidate for thelytoky . We detect a CSS peak in the intron of GB54634 ( LOC725260; Drosophila orthologue: CG9896 ) on chromosome 7 ( locus E ) . This gene encodes a protein that is currently uncharacterized across the animal kingdom: searches with NCBI BLASTp and in the Conserved Domain Database did not identify any functionally annotated orthologues or known domains . Locus F encompasses accession GB43519 ( LOC411614 ) on chromosome 11 , corresponding to the Neuropeptide Y-like ( NPY; the insect orthologue is NPF ) receptor snpfR . NPF is a neurohormone that regulates feeding motivation and behaviors in animals and is up-regulated in honeybee foragers compared to nurse bees [61] . We hypothesize that this gene is involved in lack of foraging in the Cape bee that occurs during social parasitism . On the same chromosome ( locus G ) , we also find outlier SNPs across accession GB44980 ( LOC409260 ) , which codes for an “epidermal retinol dehydrogenase 2-like” protein . Retinol dehydrogenases oxidate retinol into retinal and are important for vitamin A metabolism and the production of photopigments and retinoic acid , essential for vision , neural development , organogenesis and reproduction across animal taxa [62–64] . Locus J is centered on GB49919 ( LOC724687 ) on chromosome 13 , which codes for a protein with a Guanylate-kinase-associated protein domain ( GKAP ) and shows high similarity towards insect sequences annotated as disks large-associated protein 5-like ( DLGAP5 ) . The Drosophila orthologue is MARS ( FBgn0033845 ) , which is the fly version of the vertebrate DLGAP5-gene ( syn . HURP ) . In both the fly and vertebrates , this protein promotes microtubular stability in mitotic and meiotic spindles prior to cell division [65–68] , making it a good candidate for involvement in thelytoky in the Cape bee . Close to the end of chromosome 16 , we detect a clear selection signature across accession GB45973 ( locus L ) . This gene codes for the Dopa decarboxylase ( Ddc ) enzyme , which converts L-dopa to dopamine , a neurotransmitter with diverse functions in insects , including roles in aversive learning and memory formation , innate immunity , and ovary development [69–71] . We next analyzed patterns of genetic variation in the 12 putative selective sweep regions ( Table 2 ) . Cape bee haplotypes with strongly selected alleles can be expected to have reduced levels of genetic diversity compared to unselected Cape and African haplotypes . In a “hard” selective sweep , a single selected haplotype rises in frequency in the population , effectively reducing diversity to zero around the selected variant . In a “soft” sweep on the other hand , the selected variant already occurs on several haplotype backgrounds when selection starts , or recombine early during the selection process , such that a mix of haplotypes carrying the putative variant rise in frequency in parallel , essentially preserving more genetic variation around the selected allele compared to a hard sweep [72] . To address whether gene candidates may typically have been undergoing “hard” or “soft” selection in the Cape bees , we defined a core SNP with the highest FST value for each region . We then defined a core haplotype region as being within 0 . 01 cM of this selected SNP and measured haplotype diversity linked to the selected and non-selected allele . In every case we observed distinctly reduced genetic variation among the core haplotypes with the selected allele ( π was reduced by more than 50% in all cases ) but there was no evidence of a single core haplotype ( i . e . with no genetic variation ) being linked to the core SNP in any sweep ( Table 2 ) . This indicates that selection in these regions has most likely occurred on standing variation , and that the targets of selection were initially present on more than one haplotype , representing soft sweeps rather than hard sweeps . The patterns of genetic variation are influenced by a number of factors , including the timing and strength of selection , and the diversity of haplotypes bearing the selected variant at the onset of selection . However , it is clear that the sweeps B , D and H , that contain the strongest CSS signals , all span relatively large regions of the genome ( >10 kb ) indicating that selection may have been more recent in these regions , whereas sweeps J and L , centered around the MARS and Ddc genes , respectively , define small regions ( <6kbp ) which may represent one or more older rounds of selection . A single locus on chromosome 13 was previously implicated in control of thelytoky [17 , 18] , and it was hypothesized that a 9 bp deletion ( the thelytoky associated element 1; tae1 ) affecting alternative splicing of the gemini transcription factor within this locus governs the switch to social parasitism in the Cape bee [19] . However , a recent study did not find any association with thelytoky and the chromosome 13 locus , and found that the 9 bp deletion was polymorphic in populations of other honeybee subspecies [20] . It has also been suggested that this locus could be associated with reproductive dominance shown by Cape bee workers [18–20] . In the current honeybee genome build ( OGSv3 . 2 ) the gemini locus maps to adjacent gene accessions , GB48238 and GB48239 , which lie within a 10 kbp region on chromosome 13 . The average CSS values of these accessions are both <0 . 3 , which is below the average of all CSS values in the dataset ( CSSmean = 0 . 383 ) . None of the SNPs in this region are present among our candidates for selection . It is therefore unlikely that a major locus for thelytoky or any other Cape bee trait lies within this region . We also used the mapping of short reads to investigate the segregation of the 9 bp deletion in our dataset ( S9 Fig ) . The honeybee reference genome sequence does not contain the deletion . In some of our samples , at least one read mapped across the deletion locus with an alignment gap , which is strong evidence for the presence of a deletion in at least one allele . In other samples , at least one read mapped across the deletion without such a gap , which is evidence for the presence of at least one non-deletion allele . Finally , there are some samples where no reads map across the locus , which suggests that a deletion is likely present , although considering the relatively low average read depth , such patterns could also occur by chance . We find similar proportions of these read mappings in all three African subspecies , and we therefore estimate that the frequency of the deletion is similar ( 30–40% ) across these populations . Importantly , we find strong evidence of the presence of the non-deletion allele at high frequencies in the Cape bee samples , and the presence of the deletion allele in both A . m . scutellata and A . m . adansonii . It is therefore not possible that this deletion controls the switch to social parasitism in the Cape bee .
Cape bees ( A . m . capensis ) inhabit the biodiverse coastal Fynbos ecoregion of South Africa , whereas the neighboring savannah habitats of the Central Plateau are inhabited by the closely related A . m . scutellata ( Fig 1 ) . In its natural habitat , Cape bee worker reproduction is rare in colonies with a queen [73] , but increases if the queen is lost or during reproductive swarming [12] . Nevertheless , worker reproduction contributes significantly to the Cape bee population [74] . Cape bee workers can also act as parasites , by entering colonies of A . m . scutellata and laying eggs , even in the presence of the queen , which is eventually lost [75] . The reproducing Cape workers do not participate in foraging [76] and the colony gradually dwindles and finally dies [77 , 78] , but the offspring of the parasitic workers is able to disperse and parasitize other nests . This dramatic invasive behavior has been coined social parasitism . A number of traits observed in Cape bees facilitate this behavior , including development of worker ovaries , reproductive dominance and reproduction by thelytokous parthenogenesis . The reason why these traits appear in the Cape bee is unclear , but it is possible that they have an adaptive advantage in the Fynbos area where they predominantly occur ( reviewed in [5] ) . In addition to biological interest , the social parasitism can have disastrous impacts on apiculture as shown during the “capensis calamity” in the early 1990s when cape bees were introduced outside of native range and had severe negative effects on the native colonies of A . m . scutellata [78–80] . We hypothesized that the traits involved in social parasitism in the Cape bee represent derived adaptations that have been produced by positive selection in the Cape bee population . We analyzed genetic variants with high FST in the Cape bee population compared to a background population consisting of two other African subspecies . We find that SNPs with high FST in this comparison are strongly associated with signals of selection in the Cape bee population inferred by the PBS and XP-EHH tests , rather than being associated with selection in the background population . In the reciprocal comparisons analysing highly differentiated SNPs in the two other African subspecies , similar strong associations with selection signals were not observed . This indicates that positive selection had been more prevalent in the Cape bee population , and is therefore likely to be associated with the unique set of derived traits . To identify genetic variants that control these traits we performed a genome scan for loci under selection using the CSS statistic that combines FST and XP-EHH . We find clear signals implicating multiple loci in the suite of traits specific to the Cape bee . Selection signals are found in several genomic regions and encompass a diverse variety of genes , consistent with the diverse set of traits connected to social parasitism . Notably however , we find no evidence for a connection between a putative thelytoky locus on chromosome 13 and social parasitism in the Cape bee [18] , as there are no SNPs in this region with high FST or CSS in our dataset . In addition , we find that a 9 bp deletion , previously suggested to be associated with thelytoky in the Cape bee [19] , likely segregates at similar frequencies in other African populations , as also found in a recent study [20] . It is therefore highly unlikely that genetic variation in this region , in particular the 9 bp deletion , has any connection to social parasitism in Cape bees . Our results suggest that multiple loci , rather than a single master regulator , are responsible for the switch to social parasitism in Cape bees . The candidate loci presented here can be used in further experiments to elucidate the biological basis of these traits . Some previous studies have used F1 hybrid queens from Cape bees and another subspecies backcrossed to Cape bee drones to analyze the segregation of the putative thelytoky locus [18–20] . Laying workers produced from such crosses exhibited a ratio of thelytoky:arrhenotoky close to 1:1 predicted if thelytoky was controlled by a single locus . However , it is important to note that while this is consistent with control by a single locus , it could still be observed if the hybrid was heterozygous at multiple loci that control thelytoky . In addition , there are a number of additional traits exhibited by the Cape bee that facilitate worker reproduction and social parasitism but are not functionally connected to cell division , which are likely to be controlled by different loci . It has previously been found that ovary activation and pheromone excretion only co-vary in Cape bees but are not fully associated ( some individuals express only one trait ) , suggesting that they are controlled by more than one locus [81] . Genetic markers that co-vary with the Cape bee phenotypes and distinguish Cape bees from the scutellata population have previously been lacking or found to be inconsistent [3] . The highly differentiated SNPs identified here also have the practical benefit of being able to unambiguously identify the degree of introgression from Cape bees using only a few SNP markers . For example , there are two regions , on chromosomes 1 and 11 , respectively , with fixed variants that separate all Cape bees in our samples from all other African bees in our sample . Typing these two genetic markers would unambiguously distinguish Cape bees from neighboring populations and allow confident assessment of gene flow involving Cape bees . The Cape bees from our two sample locations have different degrees of differentiation compared to the other African bees: there are seven times as many fixed variants between the western Cape Town sample and other African bees as there are between the eastern Port Elizabeth sample and other African bees in our dataset ( 423 variants across 10 chromosomes vs 60 variants across 2 chromosomes , respectively ) and the haplotypes on which these fixations occur are more divergent from other African bees in the Cape Town sample . Stronger selection pressure or higher degrees of reproductive isolation could be driving these differences . The western Fynbos region can be considered at the current center of the Cape bee genotype , whereas the eastern subpopulation , located closer to the transitional zone , may have been subject to more introgression from scutellata-type genotypes . This is supported by a western-to-eastern cline observed in phenotypic variation [3]: worker bees from the western region have larger ovaries with up 50% more ovarioles than do the eastern bees [22 , 82] , they produce diploid female offspring through thelytoky at higher ratios [83] and although they do not mate , they have more developed spermathecae compared to the eastern bees [82] , rendering them more queen-like . Cape bee workers can lay diploid unfertilized eggs through thelytokous parthenogenesis . At the cytological level , the diploid nucleus of the egg is produced after the second meiosis . Following an atypical linear orientation of the two meiotic spindles , four nuclei are produced and the central pronucleus fuses with the central descendant of the first polar body , migrates into the egg and initiates normal embryological cleavage as if the egg had been fertilized by a sperm [6] . Among our outlier SNPs are non-synonymous variants in three candidate genes , which code for proteins that could potentially play roles in chromosomal segregation and thelytoky in the Cape bees . These genes modulate the interaction between the centrosome ( the microtubule organizing center of the cell ) and the meiotic spindle ( the structure responsible for separating chromatids during cell division ) . The gene GB45239 at locus H codes for an uncharacterized protein that has peptide sequence similarity towards an SMC domain . SMC proteins bind chromosomes and are essential for organizing the segregation of sister chromatids during cell division [84 , 85] . The gene GB50742 in locus C codes for a BBS1-like protein . BBS proteins are typically associated with the assembly of the primary cilium . Mutations in BBS genes are implied in many ciliopathic disorders and have been shown to disrupt the establishment of planar cell polarity [86 , 87] . Some BBS proteins , including BBS1 , are also localized to the centrosome and their knockouts and mutants adversely affect spindle microtubule and pole function , resulting in misaligned chromosomes during mitosis in mice [88] . The gene GB49919 ( locus J ) gene codes for a protein with a GKAP domain and the Drosophila orthologue MARS has been shown to be required for successful centrosomal binding and the maintenance of spindle bipolarity during chromosome segregation [65–68] . In Drosophila , a single point mutation in the structurally similar guanylate kinase enzyme has been demonstrated to convert it into a spindle orientation protein [89] , suggesting that the spindle apparatus is sensitive even to small genetic changes . Cape bee worker development is unusually fast , with workers emerging about a day earlier than workers of European and other African subspecies [90 , 91] . Cape bee workers also have well-developed and readily activated ovaries [3] . Morphogenesis and ovary development and activation in honeybees are associated with changes in juvenile hormone ( JH ) and ecdysteroid levels [34–36] . Ecdysteroids are also produced in ovaries , deposited in eggs and regulate embryogenesis [42] , including gonadal development in the larva [92] . In queen-destined honeybee larvae , increased levels of JH promote ovary development by protecting ovarian tissues from apoptosis [58 , 93] . We detect signals of selection across several loci that may modulate levels of these hormones , including a cluster of glucose-methanol-choline ( GMC ) oxidoreductases located on chromosome 1 . In particular the strongest signal in our dataset is found in the ecdysis triggering hormone receptor ( Ethr ) . Ecdysis triggering hormone is essential for initiating ecdysis among insects , and directly regulates JH synthesis in mosquitos [38 , 39] . The SDR11-like locus ( locus A ) shows high peptide similarity towards farnesol dehydrogenase ( FOHSDR ) in other bees and insects , which is responsible for converting farnesol into farnesal , an early step in the pathway for JH biosynthesis . Worker ovary size and JH signaling both correlate with worker behavior [94] . We hypothesize that these loci modulate larval development , ovary development , and/or worker behavior through their effects on JH and ecdysteroid levels . Our scan identifies a locus coding for a 4-coumarate-CoA ligase . 4-coumaric acid is present in jelly fed to worker-destined larvae and triggers a genetic cascade resulting in worker caste determination and reduced ovary development [95] . It is therefore possible that variants in this gene in the Cape bee may promote ovary activation in workers . Among the top signals is a sweep localized to the gene coding for dopamine decarboxylase ( Ddc ) that include several changes to the peptide sequence . Dopamine levels are associated with both memory formation [96] , behavior and ovary development in workers and is normally controlled by queen pheromones [97] . Royal jelly fed to queen-destined larvae is rich in dopamine [98] . When fed to workers , royal jelly stimulates production of dopamine and other neurohormones in the brain , induces ovary activation and reduces the willingness of workers to participate in foraging [99] . Under queen-less conditions , dopamine levels as well as gene expression for biogenic amine receptors are upregulated in workers and correlated with ovary activation and reproductive status [71 , 100 , 101] . This locus has thus been annotated for functions and effects that closely mirror the Cape bee phenotype . Our analyses for selection therefore detects candidates that can be linked to three important aspects of the Cape bee phenotype: i ) loci involved in centrosome function may affect chromosomal segregation and be responsible for Cape bee thelytoky; ii ) loci involved in ecdysteroid , JH and dopamine signaling may promote worker ovary development; iii ) loci involved in worker behavior including reproductive dominance and foraging . Given the number and nature of selection signals and candidate loci that are revealed in our comparative genome scan , we reject a model that explain all Cape bee phenotypes as being controlled by a single master-switch locus , and instead propose that a mix of multiple genotypes affect different phenotypes , which may have evolved independently .
Genome variation in the A . m . capensis ( Cape bee ) population was inferred by sequencing diploid worker bees sampled from different colonies at two locations in the coastal Fynbos ecoregion of South Africa: five bees each were sampled from the western ( Stellenbosch , Cape Town ) and eastern parts ( Kragga Kamma Game reserve region; Port Elizabeth; Fig 1 ) , respectively , for a total chromosome sample size of 20 . In both of these locations , laying workers produce nearly exclusively female offspring . The documented ratio of male:female offspring is 0:1 in these locations [22] . Additional worker bees sampled from two sub-Saharan honeybee subspecies without the social parasite phenotype were taken as representatives of a unselected African background population ( n = 40 ) : ten A . m . scutellata worker bees were collected from apiaries in the Pretoria region and ten A . m . adansonii worker bees were sampled from apiaries in Kaduna state , Nigeria . In addition , 28 European worker bees collected in Spain ( A . m . iberiensis; n = 9 ) and Scandinavia ( A . m . mellifera; n = 19 ) were used as a distantly related European outgroup . Short reads were mapped against version 4 . 5 ( Amel_4 . 5 ) of the honeybee genome [102] and the African SNP dataset included 6 . 2 million phased genotypes . The samples were originally sequenced as part of a global survey of genetic variation . Refer to Wallberg et al . [21] for additional information about collection sites and the mapping and genotyping pipelines . Cape bee workers rarely produce haploid males but frequently produce females [25] . Although the ability to produce female offspring by thelytokous parthenogenesis and act as a social parasite is a trait that appears to be fixed ( or close to fixation ) in Cape bee populations , most reproduction in such populations is still due to sexual reproduction by queens . The exception to this is when the queen is lost or during reproductive swarming , and a new sexually reproducing queen may be produced by an unfertilized worker-laid egg . We therefore expect that patterns of variation and linkage disequilibrium in Cape bees should be similar to populations of other subspecies , which is observed . Overall levels of genetic variation are similar between Cape bees and other African subspecies [21] , suggesting that the presence of asexual reproduction does not markedly reduce the effective population size in the Cape population . Episodes of positive selection should therefore leave characteristic patterns of selective sweeps in restricted genomic regions typical of an outcrossing , sexually reproducing species . We used a combination of allele frequency and haplotype structure analyses of the Cape bee genome to identify candidate variants and genes that occur in genomic regions with signatures of selection . Allele frequency differences between the Cape bees and the African background population ( scutellata + adansonii ) were calculated to estimate the per-SNP ( n = 6 , 245 , 176 ) fixation index FST using the Weir-Cockerham equation [103] and identify fixed or nearly-fixed variants between the two populations . Genetic divergence between the Cape bees and the African reference populations were inferred across the full genome and over non-overlapping 1kbp and 100kbp windows using the FST estimator of Reynolds et al . [104] . The window-based FST estimates were then transformed into divergence times T and used to estimate the Population Branch Statistic ( PBS ) , according to the procedure described by Yi et al . [27] . By comparing the divergence times between the two African populations to the corresponding divergence times between each population and the European outgroup , the PBS makes it possible to identify the divergent regions in which genetic changes are associated specifically with drift or selection in the Cape bees . The PBS therefore was computed from three pairwise FST comparisons: i ) Cape bees vs the African background population; ii ) Cape bees vs the European outgroup; and iii ) the African background population vs the European outgroup . PBS>0 identifies genomic regions in which the Cape bees diverge from both the SA population and the European outgroup , ie uniquely derived alleles in the Cape bees . To detect changes in haplotype diversity around candidate loci in the Cape bees , we applied the cross-population extended haplotype homozygosity ( XP-EHH ) test as implemented in the program selscan [28] for every SNP with a minor allele frequency ( MAF ) > = 0 . 02 . While strong selection is expected to generate long haplotypes of linked variants , recombination should counter-act to disassociate them and degrade the haplotypes over time . A recombination map previously inferred from the African honeybee genotypes [105] was used to compute the average recombination rates over windows of 100kbp . These rates were next used to specify genetic distances between the SNPs and incorporated into the EHH analyses . Honeybee recombination rates are extremely high , on average ~25cM/Mb , making them the highest rates recorded in any animal species [105–107] . We therefore expect haplotype homozygosity patterns to decay relatively quickly around selected loci , allowing for high precision in the haplotype scan . The decay of EHH in the Cape bees were queried against the corresponding patterns observed in the African background population at each SNP and the output was normalized . The program traces the decay of haplotype homozygosity around every putative “core” SNP . The average SNP density in the dataset is ~1 SNP per 30bp and estimates for core SNPs near long gaps to the next variable position ( >10 , 000bp ) that had not yet decayed below the EHH cutoff threshold ( 0 . 05 ) were discarded , to avoid inference in regions with unusually little data or uncertainty about the physical distance between markers . Using this filter , XP-EHH estimates for 0 . 35% of SNPs discarded , the vast majority of which were located close to scaffold borders . In addition to per-SNP XP-EHH scores , we computed the average score across the same windows for which we had previously estimated divergence and the PBS . In our framework , the most promising candidate loci are those that contain FST outlier SNPs associated with unusually high PBS and XP-EHH scores . In order to identify the SNPs and genes with the most comprehensive evidence of selection , we computed a single unbiased Composite Selection Score ( CSS ) by combining FST and XP-EHH estimates , following the procedure proposed by Randhawa et al . [24] . To compute the CSS , all SNPs are ranked for each statistic . For each SNP , the fractional rank positions for FST and XP-EHH , respectively , are converted into two z-statistics , the mean of which corresponds to a single joint rank for both statistics . The corresponding p-value is retrieved from a standard normal distribution . The CSS score is then taken as–log10p . We also computed CSS scores across 1kbp and 100kbp non-overlapping windows , by taking the Reynolds et al . [104] FST estimator and the average XP-EHH estimated across each window . The windows were ranked and assigned a single CSS score using the same approach as for the SNP dataset . For full-gene estimates , we then took the mean CSS of the 1kbp windows overlapping each gene body , including 2kbp of neighboring sequence to either side of the start/stop position . SNPs were associated with genes using coordinates provided in the latest official gene set ( OGSv3 . 2 ) [102] and the NCBI Annotation Release 102 ( AR102 ) , spanning ~13 , 000 accessions . A restricted set of candidate SNPs and genes were then evaluated for significantly enriched gene ontology ( GO ) terms that could indicate selection on specific biological pathways and functions . The mapped scaffolds of the honeybee genome are about 200×106 bp and the mean density of genes is roughly one accession per 15kbp . The mean physical gene length is about 8kbp , as is the intergenic regions in between genes on the same scaffolds . Intergenic SNPs were annotated and included in the GO and downstream gene candidate analyses only if they were within 8kbp of the closest gene ( 43% of all intergenic SNPs ) . We analyzed the Drosophila melanogaster orthologues of these gene sets using the GOrilla web platform [30] . To address whether gene candidates may have been subject to “hard” or “soft” selection in the Cape bees , we assessed the levels of genetic diversity across the twelve selective sweeps identified in our scan . These sweeps were associated with 25 gene candidates with comprehensive selection signals . For each sweep , a single core SNP was chosen from the outlier SNPs in the sweep . The core SNP was taken as the SNP with the highest FST value in the sweep region . In the case of a tie between more than one top FST SNP , the SNP with the highest XP-EHH value among the competing SNPs was designated to be the core variant . Cape bee haplotypes with the high frequency variant at the SNP were classified as putatively selected haplotypes , whereas Cape bee haplotypes without the variant ( if any ) were put together with adansonii and scutellata haplotypes as unselected haplotypes . We estimated genetic diversity across the region by computing the average pairwise differences between all haplotypes within each of the “selected” and “unselected” groups using the π statistic . π was computed for non-overlapping windows of 0 . 01cM of genetic length for up to 20cM away from either side of the core SNP ( 2000 windows to either side of the core ) . π was also computed for non-overlapping windows of 1kbp physical length for up to 2Mb away to either side of the core SNP . For every window , the relative difference in genetic diversity between the selected and unselected haplotypes was computed as: Δπ= ( πselected−πunselected ) πunselected We traced diversity in windows upstream and downstream of the core SNP . In each case , the center ( or first ) core window contained the core variant itself . At every distance interval away from the SNP , the average π was computed from the corresponding upstream and downstream window , effectively folding the two-sided sweep profile across the variant into a one-sided sweep profile . A genomic average background Δπ between the two groups , ΔπB , was computed from averaging Δπ across the 1000 most distant windows around all twelve sweeps . The two analyses trace diversity up until different maximum distances away from the core , but ΔπB was found to be close 0 in both analyses ( ΔπB = -2% across the 1kbp windows; ΔπB = 4% across the 0 . 01cM windows ) . Δπ was negative for all core windows containing the core variants but approached or passed ΔπB between the two groups within the first 30 windows regardless of using genetic or physical distance as units of distance . The length of a sweep was taken as the first distance interval at which the average Δπ was equal to or higher than ΔπB . In the Cape bee , the genetic basis of thelytoky was investigated previously using backcrosses of hybrids formed by crossing A . m . capensis with A . m . carnica [17] . Out of four laying workers produced by the A . m . carnica backcross , none were thelytokous , whereas out of 31 laying workers produced by the A . m . capensis backcross , the proportion of thelytoky to arrhenotoky was not significantly different from 1:1 . This pattern is consistent with inheritance of thelytoky being determined by a single recessive locus , termed thelytoky ( th ) by the authors of the study . However , it should be noted that such a pattern could also be produced if multiple loci were involved . A similar backcross scheme to produce workers that were hypothetically homozygous or heterozygous for the putative th locus was used for genetic mapping using a panel of microsatellites . A significant association with the mode of reproduction was found on chromosome 13 , which was inferred to be the location of the th locus [18] . The amount of queen substance was found to be higher in thelytokous workers produced by this backcross , and the onset of egg-laying to be earlier , leading the authors to suggest that th also controls these traits , which are related to reproductive dominance . A subsequent study used RNAi to knock out an exon of the candidate gene gemini within this region , and showed that it resulted in worker ovary activation , implicating it as the th locus [19] . It was proposed that a 9 bp deletion in the intron of the gemini gene regulates alternative splicing , thus generating thelytoky and its associated traits . More recently , however , Chapman et al . [20] were unable to replicate these results . Using reciprocal backcrosses of A . m . capensis and the closely related subspecies A . m . scutellata , they found no evidence of an association between observations of thelytoky and a genetic marker linked to the th locus . They also found that the 9 bp deletion was common in other populations of A . mellifera without any reported thelytoky . This observation indicates that it is unlikely that the 9 bp deletion determines thelytoky in Cape bees . The genetic control thelytoky in the Cape bee is therefore far from clear from these studies . For instance , although backcrosses produce patterns of segregation consistent with a single locus controlling thelytoky , it is possible that more loci are involved in this trait . | Honeybee colonies are mainly composed of female worker bees that are unable to lay eggs and reproduce; a task that is performed by a single queen bee . However , in a subspecies of the Western honeybee ( Apis mellifera ) known as the Cape bee ( A . m . capensis ) , worker bees can lay eggs produced by an abnormal form of meiotic cell division known as thelytoky . These eggs are produced asexually and develop into female worker bees , which are then able to reproduce in the same way and hence to propagate this trait . Cape bee workers also display a variety of characteristics that enable them to invade foreign colonies , reproduce and feed off their resources , a behavior known as social parasitism . The genetic basis of these traits is unknown . Here we compare whole genome sequences of Cape bees with other honeybee subspecies and identify the genes that likely control this reproductive strategy . Although Cape bees are extremely genetically similar to other African bees , we identify genetic variants that make it possible to unambiguously distinguish them . In contrast to previous studies , our results suggest that multiple genes are involved in social parasitism , including those involved in hormonal signaling that may cause worker bees to activate their ovaries and those involved in chromosomal segregation , which may cause the switch between normal meiosis and thelytoky . Our results give insights into how genetic variants that facilitate asexual reproduction are able to invade a sexually reproducing population . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"invertebrates",
"ecology",
"and",
"environmental",
"sciences",
"honey",
"bees",
"population",
"genetics",
"invertebrate",
"genomics",
"animals",
"population",
"biology",
"bees",
"genomic",
"signal",
"processing",
"hymenoptera",
"parasitism",
"genetic",
"loci",
"animal",
"genomics",
"insects",
"arthropoda",
"community",
"ecology",
"signal",
"transduction",
"haplotypes",
"cell",
"biology",
"ecology",
"trophic",
"interactions",
"genetics",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"genomics",
"evolutionary",
"biology",
"cell",
"signaling",
"organisms"
] | 2016 | Identification of Multiple Loci Associated with Social Parasitism in Honeybees |
Nipah virus targets human endothelial cells via NiV-F and NiV-G envelope glycoproteins , resulting in endothelial syncytia formation and vascular compromise . Endothelial cells respond to viral infection by releasing innate immune effectors , including galectins , which are secreted proteins that bind to specific glycan ligands on cell surface glycoproteins . We demonstrate that galectin-1 reduces NiV-F mediated fusion of endothelial cells , and that endogenous galectin-1 in endothelial cells is sufficient to inhibit syncytia formation . Galectin-1 regulates NiV-F mediated cell fusion at three distinct points , including retarding maturation of nascent NiV-F , reducing NiV-F lateral mobility on the plasma membrane , and directly inhibiting the conformational change in NiV-F required for triggering fusion . Characterization of the NiV-F N-glycome showed that the critical site for galectin-1 inhibition is rich in glycan structures known to bind galectin-1 . These studies identify a unique set of mechanisms for regulating pathophysiology of NiV infection at the level of the target cell .
Nipah virus ( NiV ) is a lethal emerging virus that infects agricultural livestock and humans . In 1999–2000 , NiV infection of agricultural workers in Malaysia and Singapore resulted in a 40% mortality rate , and subsequent outbreaks in Bangladesh resulted in an average case-fatality ratio greater than 70% [1] . In humans , NiV targets endothelial and neural cells , with resulting respiratory and neurologic sequelae; patients infected with Nipah virus often succumb to acute encephalitis with accompanying multi-organ failure due to systemic vasculitis . Autopsy studies of NiV victims identified virus in endothelial cells , with endothelial cell syncytia formation being a pathognomonic hallmark of NiV infection [2] , [3] , [4] . NiV , a member of a new genus of Paramyxoviridae , is an enveloped virus with two viral envelope glycoproteins , NiV-G and NiV-F , that mediate viral entry [5] . The NiV-G attachment glycoprotein binds to specific receptors , primarily ephrinB2 or alternatively ephrinB3 [6] , [7] , [8] . While ephrinB2 and ephrinB3 are expressed in neuronal cells , only ephrinB2 is highly expressed on endothelial cells [9] . The NiV-F glycoprotein mediates fusion of bound virus with target cells . After endothelial cell infection , NiV-F and NiV-G glycoproteins are expressed on the surface of infected cells . This triggers cell-cell fusion with neighboring endothelial cells ( homologous fusion ) or stromal cells ( heterologous fusion ) , resulting in endothelial syncytia formation , endothelial disruption , and subsequent hemorrhage and tissue damage . The NiV-F fusion glycoprotein , NiV-F0 , is initially expressed at the cell surface as a single glycosylated polypeptide precursor , but subsequently undergoes endocytosis and endosomal proteolytic cleavage by cathepsin L into F1 and F2 subunits that are disulfide linked to form the mature fusion protein NiV-F1/2 [10] , [11] , [12] . Mature NiV-F traffics back to the cell surface , where the protein can initiate cell-cell fusion at neutral pH when it is appropriately triggered by receptor binding to NiV-G . Our labs and others have previously demonstrated that specific N-glycans on NiV-F play important roles in regulating the extent of cell fusion [13] , [14] . Infection of endothelial cells by viruses results in the release of innate immune effectors , including galectins , a family of mammalian lectins [15] , [16] . Galectins are soluble , secreted proteins that stay associated with the cell surface by binding specific cell surface glycan ligands on specific glycoprotein receptors [17] . All galectins are multivalent or form higher order multimers , and can thus cross-link glycan ligands and the glycoprotein receptors that bear these ligands on the surface of cells . Human endothelial cells express galectin-1 , as well as galectin-3 , -8 , and -9 , and in vitro and in vivo activation of human endothelial cells increased synthesis as well as secretion and cell surface localization of galectin-1 [17] , [18] , [19] . We have previously shown that recombinant human galectin-1 inhibits cell fusion and syncytia formation caused by NiV-F [20] . The mechanism by which galectin-1 inhibits NiV-F mediated cell fusion is not well understood; however , we found that galectin-1 bound directly to NiV-F and caused NiV-F to oligomerize , suggesting that galectin-1 can cross link NiV-F on the surface of infected cells . Moreover , a single N-glycan site in NiV-F , the F3 glycosylation site , appears to be critical for galectin-1 mediated inhibition of cell fusion , as mutation of that site to abolish N-glycan addition , reduced galectin-1 binding to NiV-F and reduced the inhibitory effect of galectin-1 by 50% [20] . In the present study , we demonstrate that galectin-1 regulates NiV-F mediated fusion of endothelial cells and neural cells , the targets of NiV infection in vivo . Since endothelial cell syncytia formation is a pathognomic feature of NiV infection , we investigated the mechanisms involved in galectin-1 modulation of cell-cell fusion . We found that galectin-1 regulates cell fusion at three distinct points in the process; galectin-1 retains immature NiV-F0 on the cell surface to reduce production of the mature NiV-F fusion protein , galectin-1 reduces lateral movement of NiV-F on the plasma membrane that is required for cell-cell fusion , and galectin-1 directly inhibits the fusogenic activity of NiV-F by preventing fusion-peptide exposure and pre-hairpin intermediate ( PHI ) formation . The biological significance of our results is underscored by our demonstration that endogenous galectin-1 on endothelial cells is sufficient to reduce NiV-F mediated fusion . These studies identify a unique set of mechanisms for regulating the pathophysiology of NiV induced syncytia formation at the target cell level , and contribute to our understanding of the interaction between galectin-1 and glycoproteins of microbial pathogens .
We previously found that galectin-1 inhibits syncytia formation in Vero cells mediated by NiV-F and NiV-G [20] . In vivo , endothelial cells and neuronal cells are the main targets of Nipah virus [2] , due to the high expression of ephrinB2 and/or ephrinB3 by these cells [6] , [8] . To explore the role of galectin-1 in NiV-F and G mediated endothelial and glial cell syncytia formation , we used a heterologous overlay fusion assay . EphrinB2 and ephrinB3-negative cells ( PK13 ) were transfected with NiV-F and NiV-G and plated on a confluent layer of the indicated target cell type ( Fig . 1 ) . Cell-cell fusion could be observed as early as 45 min and plateaued after approximately 6 hrs . Addition of recombinant galectin-1 treatment to the co-cultures significantly inhibited syncytia formation in each of the target cell lines . Vero cells , human umbilical endothelial cells ( HUVECs ) , and microvascular endothelial cells ( mVECs ) all showed a 90% reduction in cell-cell fusion , and U87 glioblastoma cells showed a 60% reduction in fusion , compared to control cultures , in the presence of galectin-1 ( Fig . 1A ) . Representative images from the syncytia assays for each target cell type are shown in Fig . 1B . Nipah virus infection results in extensive damage to endothelial cells as a result of syncytia formation , culminating in multi-organ hemorrhage and death . During viral infection , endothelial cells become activated and release immune mediators , including galectin-1 [18] , [21] . We asked if endogenous endothelial galectin-1 could affect NiV-F mediated syncytia formation , using the heterologous fusion assay as in Fig . 1 . To activate HUVECs , the cells were cultured in 20% human serum for four days , a process that increased cell surface galectin-1 protein expression in these cells [21] . Galectin-1 expression on the cell surface was quantified by flow cytometry , and activated HUVECs demonstrated a consistent increase in cell surface galectin-1 compared to resting cells ( Fig . 2A ) . Activated HUVECs also showed a significant decrease ( 40–50% ) in cell-cell fusion , which correlated with the increase in cell surface galectin-1 ( Fig . 2B ) . Conversely , in order to determine if endogenous galectin-1 , even on resting HUVECs , was sufficient to affect NiV-F and G mediated syncytia formation , we reduced expression of galectin-1 in HUVECs using lentiviral vectors expressing siRNAs targeted against galectin-1 . A combination of three siRNAs reduced galectin-1 protein approximately 70% ( data not shown ) , and reduced cell surface galectin-1 approximately two-fold ( Fig . 2C ) . Reduction of endogenous galectin-1 had a dramatic effect on syncytia formation , as HUVECs treated with siRNA to galectin-1 demonstrated 2 . 5-fold increase in syncytia formation , compared to cells treated with control siRNA ( Fig . 2D ) . Infection of HUVECs with lentiviral vectors containing no siRNA , or siRNA against an irrelevant protein , had no effect on syncytia formation , which was identical to that observed in uninfected cells ( data not shown ) . To confirm that cell surface galectin-1 was responsible for the effect on syncytia formation , we added exogenous galectin-1 to HUVECs in which galectin-1 expression was decreased by siRNA . We observed the expected increase of cell surface galectin-1 ( Fig . 2C ) , as well as a decrease in syncytia formation ( Figure 2D ) . Thus , endogenous galectin-1 on the surface of endothelial cells inhibits NiV-F/G mediated syncytia formation , underscoring the biological relevance of the effect of galectin-1 on NiV mediated cell-cell fusion . Galectin-1 regulates the distribution and residence time of cell surface glycoproteins by binding to glycan branches on glycoproteins to create a cell surface lectin-glycoprotein lattice [22] . Lattice formation can have a variety of effects , including decreasing lateral mobility of glycoproteins . Lateral mobility is critical for effective cell-cell fusion mediated by NiV-F and NiV-G , as it is assumed that the F and G glycoproteins must physically separate in order to facilitate cell fusion [13] , [23] . To determine the effect of galectin-1 on the lateral mobility of NiV-F , we performed fluorescence recovery after photobleaching ( FRAP ) analysis , using GFP-tagged NiV-F [24] . NiV- FGFP was expressed on transfected Vero cells ( Fig . 3A ) and mediated cell-cell fusion when co-transfected with NiV-G ( Fig . 3B ) . Galectin-1 was also efficiently able to inhibit fusion mediated by NiV-FGFP and NiV-G ( Fig . 3C ) . To determine the effect of galectin-1 on NiV-F lateral movement , we measured the fluorescence recovery of NiV-FGFP ( lateral diffusion ) in the presence or absence of galectin-1 ( Fig . 3D ) . We observed 50–60% recovery of NiV-FGFP fluorescence within 20 sec of photobleaching in the absence of galectin-1 ( Fig . 3D ) . Addition of galectin-1 slowed the initial rate of fluorescence recovery , and also reduced overall fluorescence recovery to 25% ( Fig . 3D ) . Thus , the presence of galectin-1 retarded NiV-F lateral movement on the cell surface . The formation of galectin-glycan lattices on the cell surface has also been shown to reduce the rate of endocytosis of cell surface glycoproteins [25] , [26] . As described above , NiV-F is produced as an immature precursor , NiV-F0 , which is expressed on the cell surface and undergoes endocytosis and proteolytic processing to produce the fusion competent mature protein NiV-F1/2 . We asked if cell surface galectin-1 , in addition to reducing lateral mobility of NiV-F , altered endocytosis of NiV-F0 . Cells expressing NiV-F were biotinylated to label cell surface proteins and incubated at 37°C to allow for endocytosis in the presence or absence of galectin-1 . Following endocytosis , remaining cell surface biotin was removed by reduction with glutathione , and internalized biotinylated NiV-F was quantified in cell lysates . Addition of galectin-1 decreased the amount of internalized NiV-F by approximately 50% , compared to control-treated cells ( Fig . 4A ) . Galectin-1 also decreased the rate of NiV-F internalization by approximately 50% , ( Fig . 4A ) over the first 30 min , before reaching equilibrium . As galectin-1 inhibited endocytosis of NiV-F from the cell surface , and NiV-F endocytosis is required for proteolytic processing [11] , [27] , we investigated the effect of galectin-1 on proteolytic processing of NiV-F . Cells expressing NiV-F were pulse labeled and chased in the presence or absence of galectin-1 . As shown in Fig . 4B , galectin-1 treatment decreased the processing of NiV-F0 into NiV-F1/2 . At time 0 , there was little to no NiV-F0 cleavage , indicated by the absence of F1 and F2 bands . After four to six hrs , a fraction of NiV-F0 was cleaved into NiV-F1 and NiV-F2; however , the presence of galectin-1 significantly reduced processing of NiV-F0 as evidenced by the decrease in NiV-F1 and NiV-F2 . The amount of cleavage was quantified by comparing the amount of processed NiV-F ( F1+F2 ) to total NiV-F protein . Addition of galectin-1 decreased NiV-F processing by approximately 50% at the 6 hour time point ( Fig . 4C ) , which is approximately the amount of endocytosis reduction seen in the presence of galectin-1 ( Fig . 4A ) . Taken together , these results demonstrate that galectin-1 reduces NiV-F processing by binding to and decreasing internalization of the NiV-F0 precursor . To further explore the interaction between galectin-1 and NiV-F , we used a quantitative heterologous fusion assay [13] , [20] , which combines ephrinB2 positive cells stably expressing the T7 polymerase ( BSRT7 ) and ephrinB2 negative cells ( PK-13 ) transfected with NiV-F , NiV-G and a luciferase construct with a T7 dependent promoter; in this system , luciferase expression is dependent on fusion of the two cell types . We used this assay to ask if galectin-1 inhibition of fusion involved interaction of galectin-1 with mature NiV-F , as well as inhibition of NiV-F0 endocytosis and maturation , as seen in Fig . 4 . We found that , in this assay , addition of galectin-1 inhibited cell fusion by 80% , and galectin-1 mediated inhibition required the dimeric form of galectin-1 , as a monomeric galectin-1 , N-Gal-1 , was not effective at blocking cell fusion , while a covalently linked galectin-1 dimer ( GG ) was effective at blocking fusion ( Supplmentary Fig . S1A ) . In addition , in this assay , galectin-1 mediated inhibition of fusion was specific and carbohydrate mediated , as the inhibitory effect was abrogated by addition of lactose , a preferred glycan ligand for galectin-1 , but not by sucrose ( Supplementary Fig . S1B ) . Chlorpromazine is an endocytosis inhibitor that reduces NiV-F endocytosis , cleavage , and maturation [27]; cells treated with chlorpromazine at the time of transfection with NiV-G and NiV-F demonstrated virtually no maturation of nascent NiV-F0 ( Fig . 5A ) . We reasoned that , if galectin-1 plus chlorpromazine were added to cells already expressing mature NiV-F , any effect of galectin-1 on cell fusion would indicate that galectin-1 bound to and directly inhibited the fusogenic activity of mature NiV-F , rather than reducing maturation of NiV-F0 . Indeed , we found that addition of galectin-1 to chlorpromazine treated cells inhibited syncytia formation above the level of inhibition observed with chlorpromazine alone ( Fig . 5B ) , suggesting that galectin-1 can inhibit the function of the mature fusion protein , in addition to inhibiting endocytosis of NiV-F0 . To further demonstrate that galectin-1's inhibitory activity can be effected through mature NiV-F , we used a real-time fusion kinetics assay [28] ( Fig . 5C ) . Non-permissive effector cells ( receptor-negative ) were co-transfected with beta-lactamase , NiV-G , and NiV-F and then added to ephrinB2-expressing 293T target cells labeled with CCF2-AM dye . Cell-cell fusion was detected by analyzing the shift from green to blue fluorescence , indicating cytoplasmic mixing and beta-lactamase cleavage of CCF2-AM . In the absence of galectin-1 , cell-cell fusion plateaued at about 100 min after mixing of the cells , while addition of galectin-1 completely inhibited fusion in this assay . As we start to observe an effect of galectin-1 on cell fusion within 25 min , galectin-1 is likely affecting mature fusion protein already on the effector cell surface , rather than solely affecting maturation of nascent NiV-F0 . Thus , the ability of galectin-1 to inhibit fusion in this assay further supports a direct interaction between galectin-1 and the mature NiV-F fusion protein . To more specifically understand the effect of galectin-1 binding to mature NiV-F on the fusion process , we examined triggering of NiV-F in the presence and absence of galectin-1 . Current models of Nipah virus membrane fusion suggest that NiV-G binding to the cell surface receptor ephrinB2 triggers a conformational change in NiV-F; this triggering results in exposure of the fusion peptide in the pre-hairpin intermediate ( PHI ) . The PHI then undergoes six-helix bundle formation , which is the conformational change that physically drives fusion of opposing membranes [29] . In NiV-F , the two heptad repeat regions , HR1 and HR2 , fold next to each other during six-helix bundle formation . The HR1 region is transiently exposed during PHI formation , but before six-helix bundle formation . We have previously demonstrated [23] that a biotinylated peptide corresponding to the HR2 region can inhibit NiV-F mediated fusion by binding to the exposed HR1 region during PHI formation; HR2 peptide binding serves as a functional assay for formation of the PHI or NiV-F triggering . As Figs . 5B and 5C indicated that galectin-1 can block the fusogenic activity of mature NiV-F , we asked if galectin-1 binding to mature NiV-F could inhibit the conformational change in NiV-F ( triggering or PHI formation ) necessary to expose the fusion peptide , thereby inhibiting membrane fusion and syncytia formation . To detect triggering , wild type CHO cells or CHO cells stably expressing ephrinB2 ( CHOB2 ) were mixed with CHO cells transfected with NiV-F and G at 4°C for 1 . 5 hrs; NiV-G binding to ephrinB2 is an energy independent process . The biotinylated HR2 peptide was added to the cells in the presence or absence of galectin-1 , and fusion was induced by incubation at 37°C for an additional 1 . 5 hours , because PHI formation is an energy dependent process . Binding of the HR2 peptide indicates that triggering has occurred . We observed no triggering of NiV-F at 4°C , regardless of the presence of galectin-1 ( Fig . 5D ) . In the absence of galectin-1 , incubation of the cell mixture at 37°C resulted in NiV-F triggering as seen by an increase in HR2 peptide binding . However , addition of galectin-1 reduced HR2 peptide binding , demonstrating that galectin-1 inhibited triggering of mature NiV-F ( Fig . 5D ) . The change in HR2 peptide binding in the presence of galectin-1 is quantified in Fig . 5E . These results clearly show that galectin-1 can directly inhibit NiV-F triggering , so that this effect , in addition to inhibition of NiV-F lateral movement on the cell surface and inhibition of NiV-F0 maturation , contributes to the mechanism of galectin-1 mediated inhibition of NiV-F mediated cell fusion . The NiV-F fusion protein contains 5 consensus sites for N-glycosylation , labeled F1–F5; four of these predicted sites have been found to be glycosylated in vivo ( F2–F5 ) [13] . To establish the types of N-glycans expressed by NiV-F0 and NiV-F1 , glycomic screening was performed using MALDI-TOF mass spectrometry . The N-glycans were released by PNGase F and analyzed as their permethylated derivatives . The resulting spectra exhibit a series of singly charged sodiated molecular ions ( [M+Na]+ ) to which putative structures are assigned based on the molecular compositions and knowledge of the N-glycan biosynthetic pathway . The profile for the complete propeptide , NiV-F0 , can be seen in Fig . 6A . The assigned structures on NiV-F0 include high mannose ( m/z 1988 , 2192 , 2396; Man7–9GlcNAc2 ) , complex ( m/z 2040–3416; Fuc0–1NeuAc0–2Hex0–7HexNAc4–6 ) and hybrid structures ( m/z 1550 , 1999; Fuc1Hex4–5HexNAc2–3 ) . The complex and hybrid structures contain lactosamine ( Gal-GlcNAc ) moieties that can be recognized by galectin-1 . The MALDI-TOF mass spectrum of permethylated N-glycans from NiV-F1 is displayed in Supplementary Figure S2 . Comparison of the structures released from NiV-F0 and NiV-F1 suggests that the glycosylation sites on the NiV-F2 subunit ( F2 and F3 glycans ) are modified with larger complex structures than those on NiV-F1 subunit ( F4 and F5 glycans ) . Previous research has shown that galectin-1 preferentially binds to the F3 glycan and this contributes significantly to galectin-1 inhibition of NiV-F mediated fusion of Vero cells . The other NiV-F N-glycans can also contribute to galectin-1 binding , as shown by co-immunoprecipitation studies with the NiV-F3 mutant [20] . In order to confirm the differences in site glycosylation suggested in the comparative MS data , and , where possible , deduce which subtypes of N-glycan occur at each glycosylation site , we performed online nano-LC ES-MS and data-dependent MS/MS analyses on tryptic peptides and glycopeptides from purified NiV-F samples . A peptide containing the F3 glycosylation site ( GALEIYKNNTHDLVGDVR ) was observed at an elution time of ∼50 min ( Figure 6B , top panel ) . Carbohydrate structures were assigned by identifying neighboring ions separated by mass differences corresponding to sugar residues . At the F3 site , a total of 26 different glycan compositions were assigned . The carbohydrate structures are mostly complex and hybrid-type structures . The most intense peak assigned in Figure 6B ( top panel ) corresponds to a mono-sialylated , core-fucosylated bianntennary carbohydrate structure ( m/z 1019 . 384+ ) . Previous research has shown that alpha 2 , 6-linked sialic acid caps block galectin-1 binding while alpha 2 , 3 sialic acid caps do not ( Stowell et al . 2008 ) . Treatment of the tryptic glycopeptides prior to the nano-LC ES-MS experiment with Sialidase S ( a sialidase that specifically cleaves the alpha 2 , 3 linked sialic acid ) revealed that both alpha 2 , 3 and alpha 2 , 6-linked sialic acid are present , with 2 , 3-linked sialic acid being the more abundant ( Fig . 6B , lower panel ) . These results suggest that the F3 glycan contains abundant putative binding sites for galectin-1 . Analysis of the other glycosylation sites revealed that the F5 site carries only high mannose-type glycans ( Supplementary Fig . S3 ) , indicating that the F5 glycan is unlikely to contribute substantially to galectin-1 binding . The F4 site carries mostly complex-type structures , and the major peak at F4 is the sialylated biantennary glycan without core fucose ( Supplementary Fig . S4 ) , indicating that the F4 glycan may contribute to galectin-1 binding . We previously found that only the F3 glycan on NiV-F , but not the other N-glycans , is critical for optimal galectin-1 binding and inhibition of fusion of Vero cells [20] . Thus , we asked if the F3 glycan was also important for galectin-1 inhibition of fusion of endothelial cells and glial cells , the target cells of NiV . PK13 cells transfected with NiV-G and NiV-F or NiV-F3 , lacking the F3 glycan , were added to HUVECs and U87 cells , and syncytia formation was scored in the presence or absence of galectin-1 ( Fig . 7A ) . There was reduced inhibition of fusion by galectin-1 between PK13 cells expressing NiV-F3 and target cells , compared to cells expressing wildtype NiV-F3 . These results indicate that galectin-1 partially inhibits fusion in relevant cell types by binding to the F3 glycan on NiV-F . We also asked if the F3 glycan on NiV-F0 is critical for galectin-1 inhibition of endocytosis and maturation , as seen in Fig . 4A . The reduced endocytosis of NiV-F0 in the presence of galectin-1 was essentially abrogated in cells expressing the NiV-F3 mutant ( Fig . 7B ) ; in addition , we quantified processed NiV-F and NiV-F3 , and found that , as we saw in Fig . 4B , processing of wildtype NiV-F was reduced approximately 50% in the presence of galectin-1 , while there was <20% reduction of NiV-F3 processing in the presence of galectin-1 ( data not shown ) . Taken together these data indicate that the F3 glycan on NiV-F is critical for galectin-1 inhibition of endocytosis of immature NiV-F0 .
The galectins , a family of mammalian carbohydrate binding proteins , interact with a broad range of mammalian cell surface glycoproteins to regulate essential functions in virtually every type of cell , including neural , immune and endothelial cells . Recent reports have shown that galectins can also directly bind glycoproteins on microbial pathogens , including viruses , bacteria and fungi , and that some galectins can participate as innate immune effectors . For example , galectin-3 binding to Candida albicans is fungicidal [30] and galectin-4 and galectin-8 binding to Escherichia coli is bacteriocidal [31] . Conversely , some pathogens exploit endogenous galectins to promote infection or evade host immune responses [32] , [33] , [34] , [35] , [36] , [37] , [38] , [39] . Enveloped viruses use glycoproteins as both fusion and attachment molecules , and galectins have been shown to interact with envelope glycoproteins on HIV , HTLV and NiV [20] , [33] , [35] , [37] , [40] . Previous work from our laboratories demonstrated that both NiV-G and NiV-F glycoproteins bind galectin-1 , and that exogenous galectin-1 reduced syncytia formation triggered by NiV-F; galectin-1 inhibition of syncytia formation was specific for paramyxoviruses , as we did not observe inhibition of cell fusion induced by HTLV-2 , vaccinia and MLV [20] . As inflammation and viral infection can stimulate production of galectins by endothelial cells [15] , [16] , [18] , endogenous galectin-1 may contribute to host defense against NiV infection by mitigating the endothelial cell syncytia formation that is a hallmark of Nipah infection , as we have clearly demonstrated that endogenous galectin-1 can attenuate endothelial cell fusion in vitro ( Fig . 2 ) . While autopsy studies of patients who succumbed to NiV infection found endothelial cell syncytia in numerous organs including brain [2] , [3] , variation in galectin-1 expression among infected individuals may contribute to susceptibility or resistance to viral-induced pathophysiology and , in part , explain why some infected individuals do not progress to encephalitis . We investigated the effect of galectin-1 at each stage of the cell fusion process , and have identified three points at which galectin-1 inhibits NiV-F mediated syncytia formation . First , galectin-1 reduced lateral mobility of NiV-F on the plasma membrane ( Fig . 3 ) . Second , galectin-1 retarded endocytosis and maturation of the precursor NiV-F0 ( Fig . 4 ) . Third , galectin-1 prevented the triggering of mature NiV-F into the fusion-competent form ( Fig . 5 ) . Lateral mobility in the plasma membrane is important for NiV-F mediated cell fusion , as the physical separation of NiV-F and NiV-G is required for the conformational change in NiV-F necessary for membrane fusion [23] , [28] . Galectin-1 inhibition of the lateral movement of NiV-F could contribute to reduced ability of mature NiV-F to form the fusion competent PHI on the cell surface , in addition to the direct inhibition of PHI formation that we observed in Fig . 5 . All these effects are congruent with data from prior studies demonstrating that galectin interacts with endogenous mammalian plasma membrane glycoproteins to form galectin-glycoprotein lattices on the cell surface [22] , [41] . Formation of these lattices increases the local concentration of galectins at the cell surface , and also has direct effects on the glycoproteins in the lattice , as galectin-glycoprotein lattices can regulate the distribution , residence time , and function of glycoproteins on the plasma membrane [24] , [25] , [42] , [43] , [44] , [45] . Galectin-glycoprotein lattice formation has been shown to inhibit lateral diffusion of EGF receptors on tumor cells [24] . Galectin-glycoprotein lattices also reduce endocytosis of EGF receptors on tumor cells , Glut-2 receptors on pancreatic cells , and receptor tyrosine phosphatase beta on neural cells [24] , [25] , [26] . The present study provides the first demonstration that galectin-1 on the cell surface reduces lateral mobility and endocytosis of a viral glycoprotein , rather than an endogenous mammalian glycoprotein . Lattice formation between galectin-1 and NiV-F likely contributes to inhibition of NiV-F maturation , reduced mobility , and reduced triggering that we observed . Our findings also emphasize the specificity of the interaction between galectin-1 and unique glycans on a viral envelope glycoprotein . Glycoproteomic analysis showed that the most abundant glycan at the F3 site is a monosialylated biantennary glycan with a 2–3 linked sialic acid ( Fig . 6 ) . This structure fulfils the known requirement for a galectin-1 ligand [46] . Fig . 7 , as well as our previous report , shows that the F3 glycan on NiV-F participates in galectin-1 inhibition of syncytia formation at two distinct points; the F3 glycan appears essential for galectin-1 inhibiting maturation of NiV-F0 and also substantially contributes to the galectin-1 effect on the function of mature NiV-F . This level of specificity for a particular glycan on a glycoprotein counter-receptor is quite novel , in that previous reports on mammalian lectins interacting with viral glycoproteins emphasized simple pattern matching or binding to dense and repetitive arrays of terminal monosaccharide ligands [47] , [48] , [49] . In contrast , a significant component of the interaction between galectin-1 and NiV-F primarily involves a single glycosylation site on the viral glycoprotein . This interaction emphasizes the site-specific nature of glycan addition and subsequent modification during viral glycoprotein maturation and transport to the cell surface . As shown in Fig . 6 , there is significant heterogeneity among the different glycans attached to NiV-F . Similar microheterogeneity has been observed for the glycans attached to HIV gp120 [50] . This microheterogeneity would substantially contribute to selective interaction of viral glycoproteins with endogenous mammalian lectins , such as C-type lectins and galectins , or with antibodies that recognize specific glycans on viral glycoproteins . Interestingly , F3 glycan removal resulted in the highest levels of hyperfusiogenicity compared to the removal of other glycans on NiV-F [13] . This is consistent with F3 glycan removal reducing the inhibitory effects of endogenous galectin-1 . This report defines several molecular mechanisms by which galectin-1 binding to NiV-F interferes with maturation and function of NiV-F to reduce cell fusion , and also demonstrates that endogenous endothelial galectin-1 can influence the extent of syncytia formation . Since syncytia formation is the pathognomonic hallmark of NiV infection , we predict that , in vivo , galectin-1 may reduce pathophysiologic consequences during the course of an infection . An 11kb region containing the gene coding for galectin-1 contains 14 different single nucleotide polymorphisms ( SNP ) [51] , and this genetic variability may partially account for the range of pathophysiological consequences seen in Nipah virus infection , as SNPs in other galectins have been found to be associated with increased disease risk [52] . Our studies focus on cell-cell fusion , as syncytia formation is the event that primarily contributes to the endothelial destruction and hemorrhagic sequelae in NiV infection . However , galectins may also influence attachment of viruses to target cells [33] , [35] , [37] , [40] , although galectin mediated cell attachment is fundamentally distinct from the specific fusion inhibitory mechanisms that we have elucidated in this study . Galectins can play multiple distinct roles in complex biologic processes such as pathogen entry , replication and dissemination , so that our goal is to define all the roles played by galectin-1 during the entire course of Nipah virus infection of human host cells .
Vero cells and CHO cells ( ATCC ) were maintained in MEM alpha ( Invitrogen ) with 10% FBS ( Hyclone ) and 2mM Glutamax in 5%CO2 at 37°C . PK-13 porcine fibroblast cells and 293T cells were maintained in DMEM ( Invitrogen ) with 10% FBS ( Hyclone ) and 2mM Glutamax . BSR cells stably transfected with T7 polymerase were maintained in DMEM with 10% FBS ( Hyclone ) , 2mM Glutamax , and 0 . 5mg/ml G418 ( Sigma ) . U87 glioblastoma cells ( gift of P . Mischel , UCLA ) were maintained in DMEM with 10% FBS ( heat inactivated at 55°C for 30 min ) , 2mM Glutamax , and 50 units/ml penicillin/streptomycin . mVECs [6] and HUVECs ( BD Biosciences ) were maintained in MDCB-131 Complete media with fetal bovine serum and antibiotics ( VEC Technologies , INC . ) . Codon-optimized NiV-F and G plasmids tagged with AU1 were previously described [20] . NiV-F3 plasmid ( encoding NIV-F lacking the F3 glycan ) was previously described [13] . NiV-F-GFP plasmid was created by synthesizing a fusion gene between NiV-F and GFP using overlapping PCR . Oligonucleotide sequences were designed that flank the 5′ region of codon-optimized NiV-F and 3′region of GFP . An additional oligonucleotide that overlaps the 3′ region of NiV-F and the 5′ region of GFP was also designed which did not contain the stop codon at the end of the NiV-F ORF and included a GGG linker between the two genes . NiV-F ( 1 . 6 kb ) and GFP ( 0 . 8kb ) genes were amplified by PCR using the appropriate 5′ or 3′ oligonucleotide primers and the overlapping primer . The two PCR products were gel purified and used together as template for another PCR reaction using the original 5′ and 3′ primers . The resulting 2 . 2kb fusion gene product was subcloned into pcDNA3 ( Invitrogen ) and sequence verified . Recombinant human galectin-1 was expressed in E . coli and purified by affinity chromatography on lactosyl-Sepharose , as in [53]; in all assays , the buffer control includes 8mM dithiothreitol ( DTT ) , as galectin-1 is prepared and stored in PBS with DTT . PK-13 cells were transfected with codon-optimized , AU1-tagged NiV-F , NiV-F3 , or NiV-FGFP and HA-tagged NiV-G at 15ug per plasmid using Lipofectamine 2000 ( Invitrogen ) . Cells were cultured overnight , lifted with 5mM EDTA ( ethylenediaminetetraacetic acid ) and overlayed in the absence or presence of galectin-1 onto ephrinB2 positive cells ( Vero , U87 , mVEC , or HUVEC ) . After 2 hrs ( Vero/U87 ) or 6 hrs ( HUVEC/mVEC ) , cells were fixed with 2% paraformaldehyde ( EMS ) . After 4′ , 6′-diamidino-2-phenylindole ( DAPI ) staining ( Invitrogen ) , nuclei inside syncytia per ×100 field were counted by fluorescence microscopy as previously described [20] . HUVEC cells were incubated with or without 20µM galectin-1 for 30 min at 37°C . Cells were fixed in DTSSP ( Thermo Scientific ) at 0 . 2mg/ml for 10 min at room temperature , and quenched by addition of 100µl of 1M Tris pH 7 . 5 for 15 min at room temperature . Cells were washed with PBS and lifted with 5mM EDTA at 37°C for 10 min . Cells were stained with a rabbit galectin-1 antibody ( Strategic ) for 1 hr at 4°C . Cells were washed with PBS and stained with FITC-conjugated AffiniPure goat anti-rabbit IgG ( H+L ) antibody ( Jackson ImmunoResearch ) at 8µg/ml for 1 hour at 4°C . Cells were washed in PBS and resuspended in PBS with 1% BSA ( Gemini BioProducts ) for analysis by flow cytometry . Flow cytometric analysis of HUVEC cell surface galectin-1 was performed on a Becton-Dickinson FACScan , using CellQuest software ( Becton-Dickinson ) . 3 different siRNA constructs directed toward human galectin-1 mRNA in lentivectors ( pSIH1-H1-copGFP plasmid ) ( System Biosciences ) were packaged into VSV using pPACKH1 Lentivector Packaging Kit ( System Bioscience ) . Lentivirus was produced in 293T cells; three different viruses corresponding to the 3 different siRNAs were isolated , as well as a virus containing only the GFP infection marker , and a virus containing siRNA to an irrelevant protein , CD43 . HUVECs were plated at 1 . 4×105 cells per well in a 6 well tissue culture dish ( Corning ) . A single well of HUVECs was treated with each virus ( MOI of 3 . 3 , total MOI of 10 ) in PBS with 1% heat inactivated FBS ( Hyclone ) and cells were spinoculated at 2000 rpm at 37°C for 2 hrs . Infected cells were grown in full media for 4 days and knockdown was assessed by western blot and flow cytometry . Fluorescence Recovery After Photobleaching analysis of Vero cells transfected with NiV-FGFP plasmid was performed in 35mm glass bottom culture dishes ( MatTek ) on a 37°C heated stage . Cells were treated with 20µM galectin-1 or buffer control for 10 min prior to photobleaching . Images were acquired on a confocal microscope ( Leica SP2 1P-FCS ) with a HCX PL APO 63 . 0×1 . 40 oil objective and fully opened pinhole . Photobleaching of NiV-F-GFP was performed using 5 scans with the 488-nm laser at full power . Recovery data ( six cells per time point from each of two independent experiments ) was collected every 2 . 2 seconds for a total of 25 time points . Images were acquired with equivalent acquisition settings including pre-bleach , bleach , and post-bleach measurements . Bleaching and recovery were measured in a fixed area and compared to an area absent of fluorescence outside of the cell . PK-13 cells were transfected with NiV-F or NiV-F3 . After overnight incubation , cells were washed twice with KRPH buffer ( 128mM NaCl , 4 . 7mM KCl , 1 . 25mM CaCl2 , 1 . 25mM MgSO4 , 5mM Na2HPO4 , 20mM Hepes pH 7 . 4 ) and cell surface biotinylated using the EZ-Link sulfo-NHS-SS-Biotin ( Pierce ) . Biotinylation was quenched in 20mM glycine and cells were washed again in KRPH buffer . Cells were incubated in media with or without galectin-1 or buffer control for the designated times at 37°C to allow endocytosis . After timepoints , cells were washed and remaining biotin cleaved twice with cleavage buffer ( 90mM NaCl , 1 . 25mM CaCl2 , 1 . 25mM MgSO4 , 2mg/ml BSA , 50mM glutathione pH 8 . 6 ) and quenched in 20mM glycine for 15 min . Cells were lysed in lysis buffer ( 20mM Tris-HCL pH 7 . 5 , 100mM NH2SO4 , 0 . 1% BSA , 0 . 75% Trition X-100 , 0 . 01% NaN3 ) . Biotinylated NiV-F in cell lysates was quantified by ELISA using mouse anti-AU1 ( 1∶1000 ) coated Reacti-Bind goat anti-mouse plates ( Pierce ) to capture AU1-tagged NiV-F , and detected with streptavidin-HRP ( Biorad ) . 293T cells grown in 6-well plates were transfected with 0 . 25µg of NiV-F plasmid DNA and 1 . 75µg pcDNA3 per well . 24 hrs post transfection , cells were incubated in media lacking methionine and cysteine for 45 min followed by labeling with media containing 35S-cysteine and 35S-methionine ( 100mCi/ml ) for 30 min , and then in chase ( non-radioactive ) media for 4 to 6 hrs , in the presence or absence of 20µM galectin-1 . Cells were lysed in 200µl RIPA buffer ( 20mM Tris-HCl pH 7 . 4 , 137mM NaCl , 0 . 1% SDS , 0 . 5% deoxycholate , 1% NP-40 , 2mM EDTA ) . NiV-F was immunoprecipitated from cleared cell lysates using a combination of anti-NiV-F polysera [13] and anti-AU1 antibody each at 1∶100 dilution , and Protein G agarose ( Pierce ) . Precipitates were washed twice in RIPA buffer and twice with RIPA buffer plus 0 . 5M NaCl . Samples were separated on a 14% polyacrylamide gel and data were quantified and analyzed using a phosphoimager ( Molecular Dynamics 445SI ) and ImageQuant ( v5 . 2 ) . Fusion-nonpermissive PK-13 target cells were transfected with codon optimized NiV-F , NiV-G , and a plasmid containing a T7 promoter driven luciferase at a 3∶3∶1 ratio with 30µg of total plasmid DNA and grown overnight . BSR cells stably transfected with a T7 polymerase ( BSRT7 ) were lifted with 5mM EDTA at 37°C for 10 min . BSRT7 cells were co-cultured with transfected PK-13 cells for 6 hrs with or without galectin-1 . After 6 hrs , the cells were lysed in 0 . 3% Triton-X 100 by two rounds of freeze/thaw at −80°C . Luciferase expression was quantified using a Luciferase assay system ( Promega ) . Briefly , lysates were transferred to a 96 well opaque black plate , luciferase assay substrate was added , and light production was measured by luminometry ( Turner Biosystems ) . PK-13 cells were transfected with NiV-F and simultaneously treated with chlorpromazine ( Sigma ) for 16 hrs . Cells were lysed in 50mM Tris-HCl ( pH 7 . 4 ) , 1% Nonidet P-40 , 5mM EDTA , 150mM NaCl , 1mM PMSF , 10mg/ml aprotinin , 10mg/ml leupeptin , and 10mM sodium orthovanadate with scraping . Lysates were microfuged for 15 min at 10 , 000 rpm . Samples were denatured in NuPAGE reducing agent and NuPAGE Sample Buffer ( Invitrogen ) before loading . Lysates ( 10 µg ) were separated on a 12% Bis-Tris gel ( Invitrogen NuPAGE Electrophoresis System ) and electroblotted onto nitrocellulose ( Whatman ) . The membrane was blocked and probed as previously described [53] using an AU1 antibody ( Covance ) and NiV-F proteins were visualized by ECL . NiV-F triggering was measured essentially as in [23] except in the presence or absence of 0 . 4 µM galectin-1GG [54] . Briefly , CHO cells were transfected with NIV-F and NiV-G expression plasmids , plus GFP expression plasmid , at a 13∶6∶1 ratio , respectively . 18 hrs post-transfection , a 1∶1 ratio of transfected cells and either CHO ( negative control ) or CHOB2 ( CHO cells transfected with ephrinB2 [23] ) cells were mixed and incubated for 2 hr at 4°C , followed by a 90 min incubation at either 4°C or 37°C in the presence of excess biotinylated HR2 peptide and in the presence or absence of 0 . 4 µM galectin-1GG , as indicated . Cells were washed with wash buffer ( 1% FBS in PBS ) , fixed in 0 . 5% paraformaldehyde in wash buffer , and washed twice with wash buffer . Biotinylated HR2 peptide bound to F was detected using streptavidin-APC ( ebioscience ) . GFP-positive cells were gated and analyzed for HR2-biotin binding . Fusion kinetics were determined in a beta-lactamase reporter cell-cell fusion assay , as previously described [13] , [28] , [55] , [56] , using a catalytically enhanced and codon optimized beta-lactamase gene [57] , [58] . Fusion-nonpermissive PK13 effector cells were co-transfected with beta-lactamase , NiV-G , and NiV-F expression plasmids , and mixed with 293T target cells labeled with CCF2-AM dye for 30 min at 4°C . Galectin-1GG ( 0 . 4 µM ) or buffer control was added and the cells moved immediately to 37°C . Cell-cell fusion was detected by analyzing the shift from green to blue fluorescence , indicating beta-lactamase cleavage of CCF2-AM . Fluorescence was quantified every 3 min with a Synergy 2 Multi-mode microplate reader ( BioTek Instruments , Winooski , VT ) . Results are expressed as the ratio of blue to green fluorescence obtained with NiV-G- and NiV-F-transfected effectors minus the background blue to green fluorescence ratios obtained with NiV-G- and empty-vector-transfected cells . Glycan analysis was performed on NiV-F protein pseudotyped onto VSV viral-like particles produced in 293T cells . Gel bands containing purified NiV-F0 and NiV-F1 were destained using 100% acetonitrile , incubated with 10mM DTT for 30 min at 56°C , followed by 55 mM iodoacetic acid for 30 min at RT . Reduced and carboxymethylated NiV-F was digested with 0 . 5µg sequencing grade-modified trypsin ( Promega , UK ) at 37°C for 14 hrs . Tryptic peptides and glycopeptides were extracted from the gel pieces by incubating with 0 . 1% trifluoroacetic acid ( TFA ) and 100% acetonitrile; the supernatant was pooled and reduced in volume on a rotary evaporator . For glycomic screening the supernatant was lyophilised before being dissolved in 200µl ammonium bicarbonate ( 50mM , pH 8 . 4 ) and incubated with 5U of Peptide-N-glycosidase F ( PNGase F ) ( Roche Applied Science , UK ) for 24 hrs at 37°C to release the N-glycans . The reaction was terminated by lyophilisation . Glycans were separated from peptides by C18 Sep-Pak purification and permethylated as previously described [59] . MALDI-TOF MS and MS/MS data on permethylated N-glycan samples were acquired in positive ion mode [M+Na]+ using a 4800 MALDI-TOF/TOF ( Applied Biosystems , UK ) mass spectrometer as previously described [60] . The MALDI data were processed using Data Explorer 4 . 9 Software ( Applied Biosystems , UK ) . Tryptic digests were analysed by nano-LC-ES-MS/MS using a reverse-phase nano-HPLC system ( Dionex ( UK ) Ltd , Camberley ) connected to a quadrupole TOF mass spectrometer ( API Q-STAR Pulsar I , Applied Biosystems , UK ) as previously described [61] . Analysis of the ES-MS and MS/MS data was aided by use of the Peptoonist algorithm [62] . Sample was dried , resuspended in 50µl of 50mM ammonium acetate ( pH 5 . 5 ) and incubated with 10mU of Sialidase S at 37°C for 14 hrs . After digestion , NiV-F peptides and desialylated glycopeptides were desalted and separated using a C18-microtrap peptide cartridge ( Presearch , Basingstoke ) . The sample was loaded directly onto the microtrap cartridge with a 25µl gastight syringe . The microtrap was first solvated with methanol , washed off with acetonitrile and conditioned with 0 . 1% TFA . The sample was loaded onto the column and washed with 0 . 1% TFA prior to eluting with 15µl 30 and 60% acetonitrile in 0 . 1% TFA , respectively . Eluted fractions were combined and dried down gently under nitrogen before online LC-ES-MS and MS/MS analysis . | Nipah virus ( NiV ) is classified as a “priority pathogen” by the NIH . NiV infection of humans results in multi-organ hemorrhage due to endothelial syncytia formation , and also causes fatal encephalitis in up to 70% of patients . As there are no effective vaccines or therapeutics for NiV , understanding the mechanism of endothelial damage by NiV is a critical goal . Our present work defines the interaction between galectin-1 , an innate immune lectin that is secreted by human endothelial cells , with the fusion glycoprotein of NiV . We demonstrate that galectin-1 can block the function of the NiV-F protein via three distinct mechanisms , and thus reduce the ability of NiV-F to cause endothelial cell-cell fusion . Importantly , in this study , we use human endothelial cells , the primary target of Nipah virus in vivo , and demonstrate that endogenous galectin-1 made by endothelial cells contributes to limiting cell-cell fusion caused by NiV-F . As endothelial syncytia formation is one of the primary pathophysiologic events in Nipah virus infection , contributing to the hemorrhagic diathesis seen in infected patients , understanding the mechanism of endothelial cell fusion and the ability of galectin-1 to ameliorate cell fusion are critical for development of new approaches to mitigate these events . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"virology/host",
"antiviral",
"responses",
"virology/emerging",
"viral",
"diseases",
"pathology/pathophysiology",
"immunology/innate",
"immunity"
] | 2010 | Endothelial Galectin-1 Binds to Specific Glycans on Nipah Virus Fusion Protein and Inhibits Maturation, Mobility, and Function to Block Syncytia Formation |
Influenza A virus ( IAV ) neuraminidase ( NA ) receptor-destroying activity and hemagglutinin ( HA ) receptor-binding affinity need to be balanced with the host receptor repertoire for optimal viral fitness . NAs of avian , but not human viruses , contain a functional 2nd sialic acid ( SIA ) -binding site ( 2SBS ) adjacent to the catalytic site , which contributes to sialidase activity against multivalent substrates . The receptor-binding specificity and potentially crucial contribution of the 2SBS to the HA-NA balance of virus particles is , however , poorly characterized . Here , we elucidated the receptor-binding specificity of the 2SBS of N2 NA and established an important role for this site in the virion HA-NA-receptor balance . NAs of H2N2/1957 pandemic virus with or without a functional 2SBS and viruses containing this NA were analysed . Avian-like N2 , with a restored 2SBS due to an amino acid substitution at position 367 , was more active than human N2 on multivalent substrates containing α2 , 3-linked SIAs , corresponding with the pronounced binding-specificity of avian-like N2 for these receptors . When introduced into human viruses , avian-like N2 gave rise to altered plaque morphology and decreased replication compared to human N2 . An opposite replication phenotype was observed when N2 was combined with avian-like HA . Specific bio-layer interferometry assays revealed a clear effect of the 2SBS on the dynamic interaction of virus particles with receptors . The absence or presence of a functional 2SBS affected virion-receptor binding and receptor cleavage required for particle movement on a receptor-coated surface and subsequent NA-dependent self-elution . The contribution of the 2SBS to virus-receptor interactions depended on the receptor-binding properties of HA and the identity of the receptors used . We conclude that the 2SBS is an important and underappreciated determinant of the HA-NA-receptor balance . The rapid loss of a functional 2SBS in pandemic viruses may have served to balance the novel host receptor-repertoire and altered receptor-binding properties of the corresponding HA protein .
Influenza A virus ( IAV ) particles contain hemagglutinin ( HA ) and neuraminidase ( NA ) glycoproteins . HA functions as a sialic acid ( SIA ) -binding and fusion protein . NA has receptor-destroying activity by cleaving SIAs from sialoglycans . The HA and NA protein functionalities are critical for host tropism , and need to be well balanced in relation to the host receptor repertoire for optimal in vivo viral fitness [1–3] . However , there is no standard assay and unit for measuring a functional balance and the precise mode by which HA- and NA-receptor interactions contribute to the balance at the molecular level remains mostly unexplored . An optimal HA-NA balance is hypothesized to allow virions to penetrate the heavily sialylated mucus layer , to attach to host cells prior to virus entry , and to be released from cells after assembly [4–7] . Aquatic birds constitute the natural reservoir of IAVs . Occasionally IAVs from birds cross the host species barrier and manage to adapt to non-avian species , including humans . The human receptor repertoire differs from avians and requires adaptations in the SIA-interacting HA and NA proteins for optimal interaction . The HA protein of avian IAVs prefers binding to terminally located SIAs linked to the penultimate galactose via an α2 , 3-linkage . Human IAVs preferentially bind to α2 , 6-linked sialosides [8–11] . Internal sugars and their linkages as well as glycan branching have been shown to determine fine specificity of HA-receptor binding [12–17] . Changes in the receptor-binding properties of the HA proteins are achieved by mutations in the receptor binding site , which have been well documented for several HA subtypes [1 , 10 , 11 , 18] . Much less is known about the adaptations in NA required to match the corresponding HA proteins . NA is a type II transmembrane protein that forms mushroom-shaped homotetramers . Tetramerization is essential for its enzymatic activity [19 , 20] . The enzyme active site is located in the globular head domain that is linked to the endodomain via a thin stalk . The active site is made up by catalytic residues that directly contact SIA and by framework residues that keep the active site in place [21 , 22] . The catalytic and the framework residues are extremely conserved between avian and human IAVs [23] . Nevertheless , although both avian and human NA proteins preferentially cleave α2 , 3-linked SIAs , human viruses appear relatively better at cleaving α2 , 6-linked SIAs [24–27] . Adjacent to the catalytic site , NA contains a 2nd SIA-binding site ( 2SBS; also referred to as hemadsorption site ) ( S1 Fig ) [28–31] . The 2SBS is made up by three loops , which contain residues that interact with SIA . Mutations in these loops in N1 , N2 and N9 affected NA binding of erythrocytes [28 , 32–35] or sialosides [26 , 33] and enzymatic cleavage of multivalent substrates [28 , 33] but not of monovalent substrates [26 , 28 , 33] . A detailed analysis of the receptor binding properties of the 2SBS of most NAs is lacking . N1 and N2 proteins bind to α2 , 3- as well as α2 , 6-linked SIAs based on binding of resialylated erythrocytes [28 , 35] whereas N1 and N9 proteins mainly bind , via their 2SBS , to α2 , 3-linked sialosides present on glycan arrays [33] or in biolayer interferometry assays [26] . Interestingly , the high conservation of SIA-contact residues in the 2SBS of avian IAV is lost in N1 and N2 of human IAVs [1 , 26 , 28 , 30] that , supposedly , all lack a functional 2SBS . For N2 of avian viruses , the conservation of the 2SBS is only lost in viruses of the H9N2 subtype , which mainly infect Galliformes species , in contrast to other N2-containing viruses , which mainly infect Non-Galliformes species [36] ( S2 Fig ) . Conservation of the SIA-contact residues in the 2SBS of N2 is also lost in canine and not restored in swine viruses , the latter of which are generally derived from human viruses ( S2 Fig ) . It is tempting to hypothesize that the loss of a functional 2SBS in pandemic viruses is part of a required adaptation of the HA-NA balance in order to deal with the altered receptor repertoire in the novel human host [1 , 28] . At first , to test this hypothesis , a detailed analysis of the contribution of ( mutations in ) the 2SBS to receptor binding and cleavage in the context of IAV particles is necessary as the interplay with HA proteins binding to either avian- or human-type receptors needs to be taken into account . We define the HA-NA balance as the balance between the activities of HA and NA in virus particles in relation to their functional receptors on cells and decoy receptors present e . g . in mucus . We have recently established novel kinetic assays based on biolayer interferometry ( BLI ) with which , in the context of virus particles , HA binding , NA cleavage and their balance can be monitored in real time using synthetic glycans and sialylated glycoproteins [37] . Multivalent IAV-receptor binding is established by multiple low affinity interactions of several HA trimers and sialosides [38 , 39] . This enables a dynamic binding mode in which individual interactions are rapidly formed and broken without causing dissociation of the virus but providing access of NA to temporarily free SIAs . Cleavage by NA results in reduced SIA-receptor density , in virus movement and ultimately in virion dissociation [37] . How fast this occurs depends on the HA-NA-receptor balance governing the dynamics of virus-glycan interactions . In the present study , we applied these novel BLI assays to study the HA-NA-receptor balance of viruses that have a single amino acid substitution in the 2SBS . We first performed a detailed analysis of the functional importance of the 2SBS in N2 for substrate binding and cleavage by comparing NA of the pandemic H2N2 virus from 1957 , containing a mutated 2nd SIA-binding site , with an avian-like NA , in which the 2nd SIA binding site was restored . Preferred binding to α2 , 3-linked sialosides was shown to result in enhanced cleavage of substrates containing these glycans . Analysis of the HA-NA-receptor balance of viruses containing these N2 proteins in combination with H3 proteins that prefer binding to avian or human-type receptors clearly demonstrated a role for the 2SBS in the complex and dynamic interplay between HA , NA and receptor , which has been largely overlooked until now . The functional importance of the 2SBS for the HA-NA-receptor balance may explain the conservation and loss of this site in avian and human IAVs , respectively .
We first analysed the receptor binding and cleavage activity of N2 NA with and without a functional 2SBS using purified recombinant soluble NA expressed in HEK293T cells [20] . In NA of A/Singapore/1/1957 ( H2N2 ) pandemic virus ( referred to as human N2 [hN2] ) one of the SIA-contact residues in the 2SBS is mutated compared to the avian consensus sequence ( S367N , S3 Fig ) . Introduction of the reciprocal mutation ( N367S ) in this NA restored the 2SBS ( referred to as avian-like N2 [aN2] ) [28] . hN2 and aN2 displayed similar specific activities when using the monovalent MUNANA [2’- ( 4-Methylumbelliferyl ) -α-D-N-acetylneuraminic acid] substrate ( Fig 1A , S4A and S4B Fig ) , indicating that mutation of the 2SBS did not affect the catalytic activity of the N2 proteins per se . Similar results were obtained previously using membrane-associated proteins [28] , indicating that the activity of the recombinant soluble proteins accurately reflects the activity of their membrane-bound counterparts as concluded earlier for N1 [20] . Cleavage of SIAs from fetuin and transferrin sialoglycoproteins was quantified by enzyme-linked lectin assay ( ELLA ) , by analysing the increase or decrease in binding of lectins depending on their binding specificities ( S4 Fig ) . ECA ( Erythrina Cristagalli lectin ) specifically binds glycans containing terminal Galα1 , 4GlcNAc corresponding to non-sialylated N-linked sugars [40] , while PNA ( peanut agglutinin ) binds to terminal Galβ1 , 3GalNAc , which generally corresponds to non-sialyated O-linked sugars [41] . NA activity thus results in increased binding of these lectins . MAL I ( Maackia Amurensis Lectin I ) and SNA ( Sambucus Nigra Lectin ) specifically bind α2 , 3- or α2 , 6-linked SIAs , respectively [42 , 43] . Binding of SNA and MAL I is decreased by NA activity . For all lectins analysed , aN2 was more active than hN2 using fetuin , containing α2 , 3- and α2 , 6-linked SIAs ( Fig 1A ) [44] . In contrast , no statistically significant difference was observed using transferrin that only contains α2 , 6-linked sialoglycans ( Fig 1A ) [45 , 46] . Plotting the specific activities of the NA proteins relative to their specific activities as determined by the fetuin-ECA combination resulted in similar activity profiles ( Fig 1B ) , which mimic those determined previously for N1 and N9 [26 , 33] . Both hN2 and aN2 preferred cleavage of α2 , 3- ( determined with fetuin-MAL I ) over α2 , 6- ( determined with fetuin-SNA ) linked SIAs ( Fig 1B ) . In agreement herewith , the specific activities were higher when determined with the fetuin-ECA than with the transferrin-ECA combination as fetuin , but not transferrin , contains α2 , 3-linked SIAs ( Fig 1B ) . These results show that an avian-like 2SBS in N2 contributes to cleavage of the sialoglycoprotein fetuin containing α2 , 3- and α2 , 6-linked SIAs . We next used BLI to study the kinetics of NA activity on a multivalent surface coated with either an avian receptor ( 3’SLNLN: NeuAcα2-3Galβ1-4GlcNAcβ1-3Galβ1-4GlcNAc ) or a human receptor ( 6’SLNLN: NeuAcα2-6Galβ1-4GlcNAcβ1-3Galβ1-4GlcNAc ) . NA activity can be directly monitored in real-time by the specific binding of the lectin ECA to terminal Galβ1-4GlcNAc glycotopes that become available upon removal of SIA by NA ( Fig 1C and 1D , red and black lines ) . Note that cleavage of the small SIA moiety is not detected directly by BLI ( Fig 1C and 1D , dashed red and black lines ) . Binding of ECA to a sensor coated with LNLN ( Galβ1-4GlcNAcβ1-3Galβ1-4GlcNAc , Fig 1C and 1D blue lines ) rapidly reaches the maximum ECA binding signal ( representing 100% de-sialylation ) assuring that ECA binding during the relatively slow accumulation of de-sialylated glycans by NA activity ( red and black lines ) reflects the cleavage kinetics of the N2 proteins . Both hN2 and aN2 more efficiently cleaved 3’SLNLN over 6’SLNLN . Especially the aN2 protein displayed much more efficient cleavage of 3’SLNLN . We conclude that restoration of the 2SBS in hN2 to the avian consensus sequence results in enhanced cleavage of substrates containing α2 , 3-linked SIAs . The increased cleavage by aN2 of substrates containing α2 , 3-linked SIAs is expected to result from specifically increased binding to α2 , 3-linked SIAs due to the presence of an avian 2SBS , although N2 proteins were reported to bind both α2 , 3- and α2 , 6-linked SIAs by using resialylated erythrocytes [28] . We observed hemagglutination for recombinant soluble aN2 but not for hN2 ( S5A Fig ) . However , no specific binding to synthetic α2 , 3- and α2 , 6-linked sialoglycans by BLI could be observed for the recombinant soluble N2 proteins , which could be due to low affinity of the 2SBS . By embedding the N2 proteins in membrane vesicles highly multivalent receptor interactions may increase receptor-binding avidity . To this end , full length N2 proteins were expressed in 293T cells . N2 virus-like particles ( VLPs ) [47] were directly harvested from the culture supernatant , while cells were treated with hypotonic and hypertonic buffers , resulting in the release of N2 protein-containing vesicles [48] . Preparations containing similar amounts of N2 , based on MUNANA activity ( Fig 1A ) were used to determine the receptor specificity of the 2SBS by BLI [26] . Negligible binding was obtained for hN2 VLPs ( Fig 2A and 2B ) or vesicles ( S5D and S5E Fig ) to α2 , 3- or α2 , 6-linked SIAs , regardless of the presence of the NA inhibitor oseltamivir carboxylate ( OC ) , which binds the NA catalytic site . In contrast , highly 3’SLNLN-specific , binding was observed for aN2 VLPs and vesicles ( Fig 2A and 2B; S5D and S5E Fig ) in the presence of OC leading to the conclusion that aN2 has much higher lectin activity than hN2 due to the presence of a functional 2SBS . The observed α2 , 3-linked SIA specificity is in agreement with the particularly enhanced cleavage of substrates containing α2 , 3-linked SIAs ( Fig 1 ) . No binding of aN2 VLPs to 3’SLNLN was observed in the absence of OC , which is likely explained by immediate self-elution of VLPs carrying active NA proteins . To examine the contribution of the 2SBS to the HA-NA balance of virus particles we examined the replication phenotype of recombinant viruses containing either aN2 or the hN2 in the background of the 1968 pandemic virus A/Hong Kong/1/68 ( H3N2 ) ( referred to as hH3aN2 and hH3hN2 ) [28] . The hH3hN2 virus , lacking a functional 2SBS , produced large and clear plaques on Vero cells ( Fig 3A and 3B , S6A and S6B Fig ) as compared to the smaller , fuzzy plaques of the hH3aN2 virus with a functional 2SBS . Staining of plaques at 48 h post infection indicated that all cells within the plaques of hH3hN2 virus were infected , whereas many non-infected cells could be observed in the hH3aN2 plaques . This could be due to the more active aN2 , which may destroy receptors on cells before the virus can enter into the cells . hH3aN2 reached lower titres than the hH3hN2 virus at 24 and 48 h post infection when Vero cells were used ( Fig 3C ) , while no significant differences were observed for replication in MDCK cells ( Fig 3D ) . Differences in cell surface sialosides and their distribution may explain differences between replication in Vero and MDCK cells . Although the sialylation patterns of MDCK and Vero cells are poorly characterized , both cell lines can be infected with human and avian IAVs and express α2 , 3- and α2 , 6-linked SIAs [49–51] . From these results we conclude that the absence or presence of a functional 2SBS in N2 may affect virus replication kinetics in a cell type-dependent manner . Using a recently established BLI-based kinetic binding assay [37] an enhanced initial binding rate to 3’SLNLN but not 6’SLNLN ( Fig 4A and 4B ) , was observed for hH3aN2 virus containing a functional 2SBS in comparison to hH3hN2 . As a result the hH3aN2 virus displayed a higher initial binding-rate ratio 3’SLNLN/6’SLNLN than hH3hN2 ( Fig 4C , red and black bars ) . Next , two recombinant soluble glycoproteins containing mainly N-linked glycans ( lysosomal-associated membrane glycoprotein 1 [LAMP1] , ca . 18 N- and 6 O-linked glycans [52 , 53] ) , or O-linked glycans ( glycophorin A , ca . 16 O- and a single N-linked glycan [54 , 55] ) were used in BLI as recently described for recombinant fetuin [37] . LAMP1 and glycophorin A mimic the presumed functional and decoy receptors found on cells ( LAMP1 ) and on mucins ( glycophorin A ) that are rich in N- or O-glycans , respectively . Analysis of the glycans on these glycoproteins by lectin binding using BLI confirmed the presence of sialylated N-linked glycans ( both α2 , 3- and α2 , 6-linked ) on both proteins , while only glycophorin A was shown to contain sialylated O-glycans ( S7 Fig ) . Again , a functional 2SBS ( present in aN2 ) contributed to virus binding ( Fig 4D and 4E ) . This contribution was larger for binding to glycophorin A than to LAMP1 as judged from the initial binding rates ( Fig 4F ) . The HA of the 1968 pandemic H3N2 virus ( referred to as hH3 ) prefers binding to terminal α2 , 6-linked SIAs [56 , 57] . The results above implicate that , besides adaptations in HA , also adaptations in the 2SBS may contribute to a specificity-switch when an avian IAV adapts to humans . We therefore studied the effect of the 2SBS in NA when combined with an avian-type HA preferring binding to α2 , 3-linked SIAs . We generated the corresponding recombinant A/Hong Kong/1/68 ( H3N2 ) viruses containing 7 amino acid substitutions in the HA ( see S8A Fig ) . These substitutions reverted the HA back to the avian consensus sequence ( referred to as avian-like aH3 ) , including the crucial substitutions Q226L and G228S , which enable HA preferential binding to avian-type receptors [56 , 57] . The resulting viruses are referred to as aH3hN2 and aH3aN2 , depending on the absence and presence of the functional 2SBS , respectively . We confirmed the receptor-binding specificities of soluble hH3 and aH3 proteins by solid phase fetuin- and transferrin-binding assays and BLI ( S8B and S8C Fig ) . As expected , aH3 displayed higher binding levels to fetuin , containing α2 , 3- and α2 , 6-linked SIAs , than hH3 , while hH3 bound better than aH3 to transferrin , which only contains α2 , 6-linked sialoglycans . BLI analysis using H3-containing vesicles obtained from cells expressing full-length versions of hH3 or aH3 confirmed the different receptor-binding properties of these H3 proteins to 3’SLNLN and 6’SLNLN ( S8D and S8E Fig ) . In contrast to viruses containing hH3 , the presence of a functional 2SBS in aN2 enhanced replication of viruses with aH3 both on Vero ( Fig 3E ) and MDCK ( Fig 3F ) cells . Differences in virus replication were smaller for MDCK than for Vero cells . We next analysed receptor-binding properties of aH3hN2 and aH3aN2 viruses using BLI . As observed before for the hH3-containing viruses ( Fig 4C; red and black bars ) , a functional 2SBS enhanced binding to 3’SLNLN but not 6’SLNLN when N2 was combined with aH3 ( Fig 4C ) . However , viruses containing aH3 displayed similar binding kinetics in the presence of OC regardless of the presence of a functional 2SBS for both LAMP1 and glycophorin A ( Fig 4G , 4H and 4I ) . From these results we conclude that a functional 2SBS site in NA contributes to virion-receptor binding in a HA- and receptor-dependent manner . The NA enzymatic activity of the different recombinant viruses with and without a functional 2SBS was analysed using the monovalent soluble substrate MUNANA , by ELLA and by BLI . The different viruses displayed a similar NA activity per particle using the monovalent soluble substrate MUNANA ( Fig 5A ) . As also the NA proteins do not differ in their MUNANA activity regardless of the presence or absence of a functional 2SBS ( Fig 1A ) , we conclude that similar amounts of NA are incorporated into virions of the four viruses . The viruses differed , however , in their specific activities when the multivalent glycoprotein fetuin was used as substrate in an ELLA ( Fig 5B ) . hH3hN2 virus was less active compared to viruses containing aN2 and/or aH3 , indicating a contribution of receptor binding via HA and the 2SBS to NA enzymatic activity in the context of virus particles . In agreement with the results obtained with the recombinant proteins ( Fig 1B ) , cleavage of α2 , 6-linked SIA found on transferrin was less efficient and did not appear to differ significantly between the different viruses ( S9 Fig ) . The ELLAs ( Fig 5B and S9 Fig ) indicate that both receptor binding via HA and the 2SBS of NA contribute to the sialidase specific activity of virus particles . These endpoint assays do not , however , elucidate the HA-NA balance of these viruses , for which kinetic BLI assays are required [37] . Preliminary experiments showed inefficient cleavage of the synthetic glycans by the recombinant viruses . Kinetic assays to determine the HA-NA balance of these viruses were therefore performed with the glycoprotein receptors ( LAMP1 or glycophorin A ) . In the absence of OC , that is , with active NA proteins , no appreciable binding of hH3-containing viruses could be detected indicating efficient receptor cleavage by NA ( Fig 5C and 5D ) . Limited binding could be detected , however , for the aH3-containing viruses in the absence of OC . The binding curve of the virus with a functional 2SBS ( aH3aN2 ) bended earlier and had a smaller area under the curve than that of the virus without a functional 2SBS ( aH3hN2 ) for both LAMP1 and glycophorin A . This bending of the curves is explained by ongoing cleavage of SIAs by viruses attached to the sensor-attached glycoproteins , resulting in release of bound virus particles [37] . The earlier bending and smaller area under the curve observed for the aH3aN2 virus is indicative of more efficient cleavage of the sensor-attached receptors by this virus than by aH3hN2 , lacking a functional 2SBS . The effect of receptor binding via NA and HA on NA activity of virions was analysed further by NA-dependent virion self-elution from a receptor-coated BLI sensor after prior binding of the virions in the presence of OC . Self-elution of IAV particles requires NA activity and self-elution is not observed when NA activity is blocked by OC [37] . After binding of the four recombinant viruses to LAMP1 and glycophorin A in the presence of OC , OC was removed by repeated short washes in Dulbecco’s phosphate buffered saline ( PBS ) with Calcium and Magnesium and virus self-elution was monitored . Clearly , viruses with aN2 proteins eluted faster from the sensors than the viruses with hN2 ( compare hN3aN2 with hN3hN2 and aH3aN2 with aH3hN2; Fig 5E and 5F ) , for both glycoprotein receptors . Of note , NA-depended self-elution of virus particles is often preceded by an apparent increase in virus binding [37] represented here as negative self-elution , particularly in the case of aH3aN2 and aH3hN2 ( Fig 5F ) . The larger negative area of self-elution for aH3hN2 reflects the reduced NA activity of this virus compared to aH3aN2 . Also the identity of HA affected the virus self-elution rate . Viruses with hH3 eluted faster than corresponding viruses with aH3 ( e . g . compare hH3aN2 with aH3aN2 ) . For hH3aN2 , self-elution was faster from glycophorin A than from LAMP1 . For aH3-containing viruses , the opposite was observed . Differences in virion self-elution observed for different HA-receptor combinations could be due the different receptor repertoires present on the two proteins ( S7 Fig ) . The results indicate that receptor binding via the 2SBS of NA contributes to enzymatic cleavage by NA in virions and to virion self-elution from a receptor-coated surface . Virion self-elution was also shown to depend on the identity of the HA and the glycoprotein receptor used . The S367N mutation in the 2SBS of N2 was rapidly obtained after emergence of the H2N2 pandemic virus in 1957 and was observed in human H2N2 viruses until 1958 . Most viruses isolated thereafter did not contain the S367N mutations but rather contained the S370L mutation , which also results in loss of a SIA-contact residue in the 370 loop ( S3 Fig ) and hemadsorption activity [28] . Both single mutations had a similar negative effect on catalytic activity of the 1957 NA [28] . These results indicate that there was not a selection against S367 per se , but rather against a functional 2SBS , which is achieved by either mutation . However , several additional mutations accumulated in time in the three loops of the 2SBS . N2 from the A/Hong Kong/68 ( H3N2 ) ( referred to as HK N2 ) contains five mutations ( S370L , N400S , N401D , W403R , P432K ) in the 2SBS compared to the avian consensus sequence ( Fig 6A ) . To analyse the contribution of the 2SBS to the enzymatic activity of these different N2 proteins , a comparative analysis of recombinant proteins and viruses using monovalent and multivalent substrates was performed . The HK N2 protein was 4–5 fold less active than hN2 both on the monovalent substrate MUNANA and the multivalent substrate fetuin . Cleavage of sialoglycans attached to transferrin was not significantly affected ( Fig 6B ) . We also compared the NA activity of recombinant H3N2 viruses only differing in their NA segment . HK H3N2 virus , containing the 1968 HK N2 protein , displayed 2-fold lower NA activity per virus particle than the hH3hN2 virus , containing the 1957 N2 protein , as determined by MUNANA cleavage ( Fig 6C ) . Similarly , the time required for 50% self-elution of virions from the multivalent receptor LAMP1 , was 2-fold longer for HK H3N2 than for hH3hN2 ( Fig 6D ) . Thus , hN2 from 1957 and 1968 HK N2 differ to a similar extent in their catalytic activity both when monovalent or multivalent substrates are used . As receptor-binding via the 2SBS only increases NA activity for multivalent , but not monovalent substrates , we conclude that these differences do not result from differences in receptor-binding by their ( non-functional ) 2SBS . Moreover , the difference observed when comparing the two recombinant proteins is similar to the difference in activity of the two NAs in the virus context .
Since the discovery of hemadsorption activity in NA 1984 [58] and the structural evidence of the 2SBS in N9 1997 [29] , only few studies have addressed 2SBS-mediated receptor binding and the functional consequences thereof for NA activity [26 , 28 , 33 , 34 , 59] . We now show that the 2SBS is an important factor in the complex interplay between HA , NA and receptors , referred to as the HA-NA-receptor balance . A functional 2SBS in N2 was shown to prefer binding to α2 , 3-linked sialosides similarly to N1 [26] and N9 [33] . In agreement herewith , it enhances catalytic activity against substrates carrying α2 , 3-linked SIAs . The contribution of the 2SBS to the HA-NA-receptor balance of virus particles was shown to be receptor- and HA protein-dependent as demonstrated by kinetic analysis of receptor-binding and -cleavage of virions using BLI . The 2SBS was shown to contribute to receptor binding also when NA was combined with a receptor-binding HA in IAV virions , as well as to cleavage of receptors by virions and to virion self-elution from a receptor-coated surface . The absence or presence of a functional 2SBS also affected virus replication in a cell type- and HA-dependent manner . Our results indicate that mutation of the 2SBS as observed in early human pandemic viruses negatively affects the catalytic activity of NA and may serve to restore the HA-NA-receptor balance of viruses carrying HA proteins with altered receptor-binding properties in relation to a novel host sialome . Conservation of the 2SBS in most avian strains , with the notable exception of H9N2 viruses , is lost in human [26 , 29 , 30 , 34] , swine and canine variants ( S2 Fig ) . Strong conservation usually reflects a critical function . It would be very interesting to investigate in depth whether a critical function for the 2SBS in avian strains , for instance related to the HA-NA-receptor balance , is not required for efficient replication and transmission of human , canine and swine strains . N2 prefers binding of α2 , 3- over α2 , 6-linked SIAs via its 2SBS . The specificity of the N2 2SBS correlates with the enhanced cleavage of substrates carrying α2 , 3-linked SIAs compared to substrates carrying only α2 , 6-linked sialosides . Of note , enhanced activity was also observed for α2 , 6-linked SIAs at least when these sialosides were linked to substrates additionally carrying α2 , 3-linked SIAs ( Fig 1A; fetuin-SNA combination ) . These results indicate that the 2SBS enhances catalytic activity by bringing sialosides on multivalent substrates close to the catalytic site and that , depending on the substrate used , the enhanced cleavage of SIAs not necessarily matches the specificity of the 2SBS . Preferred binding of avian-type receptors via its 2SBS was previously also observed for N9 [33] and N1 [26] , suggesting that this is a conserved feature for NAs of different subtypes . We cannot exclude , however , that the 2SBS of different NA subtypes may differ in their receptor-binding fine specificity , as structural differences were observed in the interactions between ligands and the 2SBS for different NA subtypes [30] . In N9 , the conserved K432 residue in the 2SBS forms a hydrogen bond with SIA [29] and mutation K432E in N1 has a large negative effect on the cleavage of multivalent substrates [26] . In contrast , several other avian NA subtypes , including N2 , contain a Q or E residue at this position , which does not form a hydrogen bond with SIA in the few available crystal structures [30] . Previously , it was shown that N1 and N2 NAs bound with similar efficiency to both avian and human type receptors SIAs [28 , 35] . This discrepancy is probably explained by the different methods used to analyse the receptor specificity of the 2SBS . In the previous reports , a red blood cell binding assay was employed , in which desialylation of erythrocytes was followed by resialylation using α2 , 3- or α2 , 6-sialyltransferases . Binding to resialylated erythrocytes might be affected by prior incomplete desialylation . Alternatively , a higher receptor density on erythrocytes compared to the BLI sensor surface might allow for binding of α2 , 6-linked SIAs . The ability of the 2SBS to bind human-type receptors to some extent is also suggested by the modestly increased or decreased cleavage of SIAs from substrates only containing α2 , 6-linked SIAs upon the introduction of mutations in the 2SBS ( this study and [26 , 28] ) . The 2SBS contributed to receptor-binding also when NA was combined with a receptor-binding HA in IAV virions . In combination with HA preferring binding to α2 , 6-linked SIAs ( hH3 ) , the 2SBS enhanced binding for all receptors analysed , except 6’SLNLN , to which the recombinant aN2 protein did not bind . Binding to glycophorin A , carrying many O-linked sugars also found on mucins , was more enhanced by the 2SBS than binding to LAMP1 , which carries mostly sialylated N-glycans . The functional significance of this difference remains to be determined . When combined with HA that prefers binding to α2 , 3-sialosides ( aH3 ) , the enhancing effect of the 2SBS was not observed for the glycoprotein receptors analysed . Thus , the contribution of NA to virion-receptor binding depends on the specificity/affinity of the corresponding HA and the receptors present . Previously it was shown that the active site of NA contributes to virion-receptor binding in case of a low-activity catalytic site [37] , a characteristic which is also appears to be displayed by recent H3N2 viruses [60 , 61] . As we now show that a functional 2SBS in NA can also contribute to virion-receptor binding , two mechanisms exist by which NA can assist in binding of virions to host cells . A complex interplay between HA , NA and receptor determines the attachment of virus particles to and release from a receptor-containing surface . This HA-NA-receptor balance can be experimentally determined using kinetic BLI assays by analysis of virus binding in the absence or presence of NA inhibitors and self-elution from different receptors ( this paper and [37] ) . The HA-NA-receptor balance determines the residence time of a virus on a sialylated surface and the speed by which it moves over this surface . We assume that an optimal balance is important for virions to efficiently pass the heavily sialylated mucus layer , while still allowing virion attachment to host cells resulting in endocytic uptake . The complexity of the HA-NA-receptor balance is exemplified by the contribution of NA to receptor binding [37] and of HA to the apparent catalytic activity of NA ( this paper ) [37 , 62] . We now show that the HA-NA-receptor balance as reflected for example in virion self-elution ( Fig 5 ) is affected by a functional 2SBS , depending on the particular HA with which NA is combined and the receptors used . Changes in the 2SBS of NA should thus be considered in the context of mutations affecting the receptor-binding site of HA and the catalytic site of NA . The 2SBS of N2 appears to accumulate more mutations than other surface exposed parts of the NA protein ( S3 Fig ) . While the 1957 N2 protein has a single substitution in the 2SBS , the 1968 N2 protein contains five mutations in this site . The accumulation of several mutations in the 2SBS was found to have no further negative effects on the enzyme-enhancing function of 2SBS as compared to a single mutation of a SIA contact residues in the 2SBS of an early pandemic virus from 1957 . Although we cannot exclude that the accumulation of mutations in the 2SBS of N2 indicates ongoing adaptation of NA to the human host or serves to restore subtle deviations in the HA-NA-receptor balance resulting from other mutations in HA and/or NA , it seems more likely that it rather results from continuous immune pressure on this site [22 , 63] in combination with loss of functional importance of the 2SBS in human viruses . An important role for the NA 2SBS in IAV replication in vivo is suggested by the conservation of this site among NA subtypes of most avian viruses , the rapid loss of this site in human pandemic viruses ( [1 , 26 , 28 , 30 , 36] and S2 Fig ) , the important role of this site in HA-NA-receptor balance ( this study ) and observations that this site affects virus replication in vitro ( [26 , 34 , 59] and this study ) . Of note , we now show that the presence or absence of a functional 2SBS affected virus replication depending on the receptor-binding properties of HA , with which NA was combined . Replication of viruses with a human or avian-like HA is enhanced by the absence or presence of a functional 2SBS , respectively , although some cell-dependent differences were observed . The absence or presence of a functional 2SBS was reported not to affect influenza viral replication in ducks [34] . However , in this latter study recombinant viruses were used containing HA from a H2N9 and NA from a H3N2 virus . This may have resulted in a mismatched HA-NA combination in which the presence of the 2SBS might be of minor influence on replication . Alternatively , the 2SBS may be important for virus transmission rather than for replication in ducks per se . Clearly , additional experiments are needed to demonstrate the importance of the 2SBS for IAV replication and transmission in vivo . Interestingly , both for H9N2 and H7N9 viruses , the well-known Q226L mutation in the receptor-binding site of HA , resulting in a shift from avian to human receptor specificity , is associated with mutations in the 2SBS that negatively affect receptor binding [33 , 36] . These avian viruses thus display a striking parallel with the changes observed in the receptor-binding sites of HA and NA of avian-origin pandemic viruses . We propose that mutations in the 2SBS of avian viruses may be indicative of an as of yet underappreciated , increased potential of avian viruses to cross the host species barrier . Of note , also upon introduction of coronavirus OC43 into humans , the lectin function of the receptor-destroying hemagglutin-esterase protein was lost through progressive accumulation of mutations resulting in reduced cleavage of multivalent substrates [64] . Thus , both coronaviruses and IAVs appear to adapt to the sialoglycome of the human respiratory tract by tuning the virion receptor-binding and cleavage functions , the latter among others by mutation of the lectin domain of the receptor-destroying NA .
Human-codon optimized cDNAs ( Genescript ) encoding the N2 ectodomain of A/Singapore/1/57 ( H2N2 ) ( GenBank accession no . AY209895 . 1; referred to as human N2 [hN2] ) and a variant thereof containing the N367S mutation ( referred to as avian-like N2 [aN2] ) were cloned into a pFRT expression plasmid ( Thermo Fisher Scientific ) in frame with sequences encoding a signal sequence derived from Gaussia luciferase , a Strep tag and a Tetrabrachion tetramerization domain , similarly as described previously [20] . The corresponding full length ( FL ) NA-coding plasmids were generated by replacement of the non-NA coding sequences by sequences encoding the NA transmembrane domain and cytoplasmic tail of N2 of A/Singapore/1/57 ( H2N2 ) . Human-codon optimized cDNAs encoding FL H3 or the H3 ectodomain of A/Hong Kong/1/68 ( H3N2 ) ( GenBank accession no . CY033001; referred to as human H3 [hH3] ) or of an variant thereof containing 7 amino acid substitutions , which revert the HA back to the avian consensus sequence [56] ( referred to as avian-like H3 [aH3] ) were cloned in pCD5 expression vectors similarly as described previously [65] . Codon optimized glycoproteins LAMP1 and glycophorin A ectodomain-encoding cDNAs ( Genescript ) were genetically fused to Fc-tag , for Protein-A based purification , and a Bap tag [66] , for binding to octet sensors , and cloned in a pCAGGs vector , similarly as described previously for fetuin [37] . NA and glycoprotein expression plasmids were transfected into HEK293T ( ATCC ) cells using polyethylenimine ( PolyScience ) [20] . An expression vector encoding BirA ligase was cotransfected with the LAMP1- and glycophorin A-coding vectors [37] . Five days post transfection , cell culture media containing soluble NA proteins and glycoproteins were harvested and purified using Strep tactin or protein A containing beads [20 , 37] . Purified NA proteins were quantified by quantitative densitometry of GelCode Blue ( Thermo Fisher Scientific ) -stained protein gels additionally containing bovine serum albumin ( BSA ) standards . The signals were imaged and analysed with an Odyssey imaging system ( LI-COR ) . HEK293T cells were transfected with full-length NA constructs to obtain membrane vesicles . To this end , cells were vesiculated as described previously [26 , 48] . VLPS and membrane vesicle preparations were purified using Capto Core 700 beads ( GE Healthcare Life Sciences ) according to the manufacturer’s instructions and as detailed previously [67] to remove proteins smaller than 700 kDa . The amount of NA protein in the VLPs and vesicle preparations was determined using the MUNANA assay described below . Generation of recombinant virus HK H3N2 , which harbours all genes from the pandemic virus A/Hong Kong/1/68 ( H3N2 ) has been described before [57] . Also the generation of hH3hN2 and hH3aN2 viruses , which carry the N2 gene of the pandemic A/Singapore/1/1957 ( H2N2 ) in the background of A/Hong Kong/1/68 ( H3N2 ) has been described before [28] . The hH3aN2 virus contains substitution N367S in the N2 protein . aH3hN2 and aH3aN2 viruses were generated as described previously [56] in the background of A/Hong Kong/1/68 ( H3N2 ) . These latter viruses carry the H3 protein of A/Hong Kong/1/68 ( H3N2 ) containing 7 amino acid substitutions in HA which revert the HA back to the avian consensus sequence [56] combined with the N2 protein of A/Singapore/1/1957 ( H2N2 ) with ( aH3aN2 ) or without ( aH3hN2 ) the N367S substitution . Virus stocks were grown in MDCK-II cells ( ECACC ) . Viruses were inactivated by UV radiation using UV Stratalinker 1800 ( Stratagene ) on 50 , 000 μJoules prior to their use in the binding and cleavage assays . UV inactivation did not affect the enzymatic activity of NA as determined with the MUNANA assay . The NA enzymatic activity was determined by using a fluorometric assay [68] in combination with 2’- ( 4-Methylumbelliferyl ) -α-D-N-acetylneuraminic acid ( MUNANA; Sigma-Aldrich ) as described previously [20] . Enzymatic activity of the NA proteins towards multivalent glycoprotein substrates was analysed using a previously described enzyme-linked lectin assay ( ELLA ) [33] . In brief , fetuin- or transferrin-coated plates were incubated with serial dilutions of recombinant soluble NA proteins . After overnight incubation at 37°C , plates were washed and incubated with either biotinylated Erythrina Cristagalli Lectin ( ECA , 1 . 25 μg/ml; Vector Laboratories ) , biotinylated peanut agglutinin ( PNA , 2 . 5 μg/ml; Galab Technologied ) , biolinylated Sambucus Nigra Lectin ( SNA , 1 . 25 μg/ml; Vector Laboratories ) or biotinylated Maackia Amurensis Lectin I ( MAL I , 2 . 5 μg/ml; Vector Laboratories ) . Cleavage of SIAs from fetuin and transferrin was quantified by analysing the increase ( PNA and ECA ) or decrease ( MAL I and SNA ) in binding of different lectins depending on their binding specificities ( S4 Fig ) . The binding of ECA , PNA , SNA and MAL I was detected using horseradish peroxidase ( HRP ) -conjugated streptavidin ( Thermo Fisher Scientific ) and tetramethylbenzidine substrate ( TMB , bioFX ) in an ELISA reader EL-808 ( BioTEK ) by measuring the optical density ( OD ) at 450 nm . The data were fitted by non-linear regression using the Prism 6 . 05 software ( GraphPad ) . The resulting curves were used to determine the amount of NA protein corresponding to half maximum MUNANA cleavage or lectin binding . The inverse of this amount is a measure of specific activity ( activity per amount of protein ) and was graphed relative to other NA proteins or substrate-lectin combinations . Plaque assays were performed in Vero cells ( ATCC ) as described previously [69] . One hour after infecting the cell monolayers with 30–50 plaque forming units of the virus in 1 ml of maintenance medium , the virus inoculum was removed and cells were covered the Avicel RC-581 overlay medium and cultures were incubated at 37°C in 5% CO2 atmosphere . After three days of incubation , the overlay was removed by suction and the cells were fixed with 10% formalin and stained with 1% crystal violet solution in 20% methanol in water . For immunostaining , cells were fixed with 4% paraformaldehyde solution for 30 min at 4°C , washed with PBS and permeabilized by incubation for 10–20 min with buffer containing 0 . 5% Triton-X-100 and 20 mM glycine in PBS . Cell layers were incubated with monoclonal antibodies specific for the influenza A virus nucleoprotein ( kindly provided by Dr . Alexander Klimov at Centers for Disease Control , USA ) for 1 hour followed by another 1 hour incubation with peroxidase-labeled anti-mouse antibodies ( DAKO , Denmark ) and 30 min incubation with precipitate-forming peroxidase substrates True Blue . Stained plates were washed with water to stop the reaction , scanned on a flatbed scanner and the data were acquired by Adobe Photoshop 7 . 0 software . To characterize replication kinetics of different recombinant viruses , two replicate cultures of Vero or MDCK cells in 12-well plates were infected with each virus at MOI 0 . 001 ( Vero cells ) or 0 . 0001 ( MDCK cells ) . Inocula were removed 1 hpi , fresh medium was added , and cultures were incubated at 37°C . Samples of culture supernatant were taken 24 , 48 and 72 hpi and stored frozen . They were titrated together using focus formation assay in MDCK cells as described previously [57] . Numbers of infected cells per well were counted for the virus dilution that produced from 30 to 300 infected cells per well and recalculated into numbers of focus forming units ( FFU ) per ml of the original undiluted virus suspensions . For the full length protein-containing vesicles and VLPs , similar amounts of NA activity , and thus NA protein , were applied in the BLI assays using the Octet RED348 ( Fortebio ) . Inactivated virus preparations were analysed using Nanoparticle Tracking Analysis ( Nanosight NS300 , Malvern ) as detailed below in order to use similar number of virus particles in the BLI assays . BLI assays were performed as described previously [37] . All experiments were carried out in Dulbecco's PBS with Calcium and Magnesium ( Lonza ) at 30°C and with sensors shaking at 1000 rpm . Streptavidin biosensors were loaded to saturation with biotinylated synthetic glycans 2 , 3-sialyl-N-acetyllactosamine-N-acetyllactosamine ( 3’SLNLN ) , 2 , 6-sialyl-N-acetyllactosamine-N-acetyllactosamine ( 6’SLNLN ) , N-acetyllactosamine-N-acetyllactosamine ( LNLN ) , LAMP1 or glycophorin A glycoproteins . Synthetic glycans were synthesized at the Department of Chemical Biology and Drug Discovery , Utrecht University , Utrecht , the Netherlands . For the NA kinetic cleavage assay , the sensors loaded with synthetic glycans were incubated in 100 μl buffer containing 4 μg recombinant soluble aN2 or hN2 in the absence or presence of 8 μg ECA . As controls , sensors were also incubated with ECA in the absence of N2 . Association of the N2 VLPs , vesicles and virus particles was analysed for 30 minutes in the absence or presence of 10 μM OC ( Roche ) . For viruses , the virus association phase in the presence of OC was followed by three 5 s washes and a dissociation phase in the absence of OC . Initial binding rates were determined similarly as previously described [37] . For lectin binding , the sensors loaded with recombinant glycoproteins were incubated with the different lectins ( 8 μg/100 μl ) for 15 minutes . NTA measurements were performed using a NanoSight NS300 instrument ( Malvern ) following the manufacturer’s instructions . The UV-inactivated virus preparations were diluted with PBS to reach a particle concentration suitable for analysis with NTA . All measurements were performed at 19°C . Per analysis , the NanoSight NS300 recorded five 60 second sample videos , which were then analysed with the Nanoparticle Tracking analysis 3 . 0 software , resulting in quantitative information on particle number and particles sizes ( S10 Fig ) . Each virus preparation was analysed twice and mean values were used . NTA measurements were validated by analysis of virus stocks quantified earlier by silver staining of viral proteins after electrophoresis on polyacrylamide gels [37] . Results obtained via both methods correlated well ( less than 25% deviation ) . | Influenza A viruses infect birds and mammals . They contain receptor-binding ( HA ) and receptor-destroying ( NA ) proteins , which are crucial determinants of host tropism and pathogenesis . It is generally accepted that the functional properties of HA and NA need to be well balanced to enable virion penetration of the receptor-rich mucus layer , binding to host cells , and release of newly assembled particles . This HA-NA-receptor balance is , however , poorly characterized resulting in part from a lack of suitable assays to measure this balance . In addition , NA is much less studied than HA . NA contains , besides its receptor-cleavage site , a 2nd receptor-binding site , which is functional in avian , but not in human viruses . We now show that this 2nd receptor-binding site prefers binding to avian-type receptors and promotes cleavage of substrates carrying this receptor . Furthermore , by using novel assays , we established an important role for this site in the HA-NA-receptor balance of virus particles as it contributes to receptor binding and cleavage by virions , the latter of which is required for virion movement and self-elution from receptors . The results may provide an explanation for the rapid loss of a functional 2nd receptor-binding site in human pandemic viruses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"chemical",
"characterization",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"vertebrates",
"viral",
"structure",
"orthomyxoviruses",
"animals",
"viruses",
"rna",
"viruses",
"research",
"and",
"analysis",
"methods",
"receptor",
"binding",
"assays",
"influenza",
"a",
"virus",
"birds",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"recombinant",
"proteins",
"binding",
"analysis",
"viral",
"replication",
"virions",
"biochemistry",
"eukaryota",
"lectins",
"influenza",
"viruses",
"virology",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"amniotes",
"organisms"
] | 2019 | The 2nd sialic acid-binding site of influenza A virus neuraminidase is an important determinant of the hemagglutinin-neuraminidase-receptor balance |
Although little is known about the role of the cystic fibrosis transmembrane regulator ( CFTR ) gene in reproductive physiology , numerous variants in this gene have been implicated in etiology of male infertility due to congenital bilateral absence of the vas deferens ( CBAVD ) . Here , we studied the fertility effects of three CBAVD–associated CFTR polymorphisms , the ( TG ) m and polyT repeat polymorphisms in intron 8 and Met470Val in exon 10 , in healthy men of European descent . Homozygosity for the Met470 allele was associated with lower birth rates , defined as the number of births per year of marriage ( P = 0 . 0029 ) . The Met470Val locus explained 4 . 36% of the phenotypic variance in birth rate , and men homozygous for the Met470 allele had 0 . 56 fewer children on average compared to Val470 carrier men . The derived Val470 allele occurs at high frequencies in non-African populations ( allele frequency = 0 . 51 in HapMap CEU ) , whereas it is very rare in African population ( Fst = 0 . 43 between HapMap CEU and YRI ) . In addition , haplotypes bearing Val470 show a lack of genetic diversity and are thus longer than haplotypes bearing Met470 ( measured by an integrated haplotype score [iHS] of −1 . 93 in HapMap CEU ) . The fraction of SNPs in the HapMap Phase2 data set with more extreme Fst and iHS measures is 0 . 003 , consistent with a selective sweep outside of Africa . The fertility advantage conferred by Val470 relative to Met470 may provide a selective mechanism for these population genetic observations .
The cystic fibrosis transmembrane conductance regulator ( CFTR; OMIM 602421 ) gene functions as a chloride channel that regulates salt and water transport across epithelial cell membranes . More than 1 , 600 mutations ( Cystic Fibrosis Mutation Database; http://www . genet . sickkids . on . ca/cftr/ ) in the CFTR gene cause cystic fibrosis ( CF; OMIM 219700 ) , an autosomal recessive disorder affecting the exocrine glands of the respiratory , digestive and reproductive systems . The clinical manifestations of CF in affected individuals vary widely , with both age at diagnosis and lethality ranging from the first year of life to the third ( and later ) decade [1] . One symptom , however , that is present in nearly all male CF patients is infertility due to congenital bilateral absence of the vas deferens ( CBAVD; OMIM 277180 ) , which results from blockage in the transport of spermatozoa from testicular tissues to the distal genital track [2] . Curiously , CBAVD is also a cause of infertility in otherwise healthy men , accounting for ∼2% of all male infertility cases . However , 80% of men with isolated CBAVD carry one or two mutations in the CFTR gene [3] , defining a primarily genital form of CF . The most common genetic cause of CBAVD is compound heterozygosity for a 5-thymidine ( 5T ) repeat allele at the 3′ splice acceptor site of intron 8 and a CF-causing mutation in the CFTR gene [3] . The length of the polyT tract within intron 8 is associated with splicing efficiency of exon 9 [4] . The shorter 5T allele , compared to the more common 7T or 9T alleles , results in under-utilization of the splice site and increased proportions of CFTR transcripts lacking exon 9 , which encode a nonfunctional protein . However , the 5T allele alone does not explain all cases of CBAVD . Other polymorphisms , including a TG repeat [ ( TG ) m] located immediately upstream of the polyT tract in intron 8 , and an amino acid changing polymorphism ( Met470Val; 1540A>G [rs213950] ) in exon 10 , have also been implicated [5] , [6] . For example , longer TG repeat alleles ( TG12 or TG13 ) alter the stability of the mRNA secondary structure and decrease exon 9 splicing efficiency , thereby increasing the penetrance of the 5T allele [7] , [8] . Likewise , a valine at a common polymorphism at amino acid 470 ( Val470 ) results in the CFTR protein to mature more quickly , but with lower activity compared to the methionine ( Met470 ) allele [8] . An association between the 5T and Val470 alleles in men with CBAVD but not in fertile controls led de Meeus et al . to suggest that the Met470Val locus acts as a modifier by increasing the penetrance of the 5T allele in CBAVD [5] . To further investigate the contribution of CFTR polymorphisms in male reproduction , we examined the effects of the intron 8 ( TG ) m and polyT variants and the Met470Val polymorphism on the variation in natural fertility in healthy men . We conducted this study in the Hutterites , a founder population of European descent [9] , [10] . The Hutterites provide many advantages for genetic studies of fertility . First , they practice a communal lifestyle that minimizes variation in socioeconomic , cultural , religious , and other factors that might affect reproductive practices . For example , contraceptive use is limited and a desire for large families is widespread . As a result , Hutterite family sizes are large ( median completed family size >10 in 1960's [11] , [12] ) and reproductive rates are among the highest observed in humans [13] . Although the overall allelic architecture in the Hutterites is similar to that of other European populations [14] , [15] , there are only two CF-causing mutations segregating in the Hutterites , ΔF508 and the more common , Hutterite-specific M1101K . We previously examined the effects of carrier status for these two mutations on family size and birth rate in nearly the same men considered in this study , but found no association with reproductive outcomes [16] . The results we report here , however , suggest a significant contribution of at least one common CFTR polymorphism to natural variation in male fertility , and provide further support for the role of this protein in normal male reproductive processes .
We genotyped 204 married Hutterite men for the Met470Val and intron 8 polyT and ( TG ) m repeat polymorphisms; allele and haplotype frequencies are shown in Table 1 and Table 2 . The CBAVD-associated 5T , TG12 , and TG13 alleles are either absent ( 5T , TG13 ) or rare ( TG12 ) in the Hutterites . On the other hand , the Val 470 allele occurs at high frequency ( frequency 0 . 29 ) . The Val allele resides on two haplotypes in the Hutterites , one common ( TG11-7T-Val470 ) and one rare ( TG12-7T-Val470 ) . To assess the effects of these polymorphisms on male fertility , we defined a measure of “birth rate” as the number of births per year of marriage in men with at least two children ( see Materials and Methods ) . Associations between the CFTR alleles and haplotypes and birth rate were examined in Hutterite men using a regression-based test designed for large complex pedigrees , and which corrects for the relatedness between all pairs of men in this study [17] . The results of the association studies are summarized in Table 3 . Homozygosity for the Met470 allele was associated with significantly lower birth rates in Hutterite men ( P = 0 . 0096; Figure 1A ) , and accounted for 4 . 59% of the residual variance ( after adjusted for covariates , see Materials and Methods ) in birth rate between males . The association remained significant when Val470 homozygotes ( N = 14 ) and heterozygotes ( N = 89 ) were combined ( Model 2 ) , consistent with a recessive effect of the Met470 allele on increased birth rates ( P = 0 . 0029; Figure 1B ) . Both models remain significant after adjusting for multiple comparisons by a conservative Bonferroni correction ( Model 1 Pc = 0 . 038 , Model 2 Pc = 0 . 012 ) . There was no significant association between alleles at the ( TG ) m locus and birth rate . A marginal association was observed between the 7T allele and increased birth rates ( 7T vs . 9T , P = 0 . 060 ) , but we attribute this to linkage disequilibrium ( LD ) with the Val470 allele ( D' = 1 . 0 , Table 2 ) . Nine men in this sample carried the M1101K mutation ( none were ΔF508 carriers ) . In all cases , the M1101K mutation was on the Met470 background . Therefore , to remove the potential confounding effects of M1101K , we repeated our analyses after excluding these nine men . The association with Met470Val remained equally significant ( Model 1 P = 0 . 0059 , Model 2 P = 0 . 0020 ) , suggesting that the observed fertility effects associated with Met470Val are not due to this pathogenic CFTR mutation . On the other hand , the association with the Met470Val locus in Hutterite men is quite robust . Figure 2A shows the cumulative distribution of the number of years from marriage to each birth by genotype . On average , Met/Met men achieve each birth in more time than men with one or two copies of the Val allele . The difference between the means of the genotype groups increases with increasing birth number , reflecting a cumulative , positive effect of the Val allele ( relative to Met/Met ) on male fertility . The average effect of homozygosity for the Met470 allele compared to carrying one or two copies of the Val470 is a decrease of 0 . 049 births per year of marriage ( Figure 1B ) . This corresponds to 0 . 56 fewer births over the course of an average reproductive period ( 11 . 5 [±5 . 0] years in this cohort ) . For example , Met470 homozygous men who are married 11 . 5 years or longer have a median of 7 children compared to 8 children in Val470 carrier men ( Wilcoxon P = 0 . 0002; Figure 2B ) . Finally , the time required to achieve 6 births ( the overall mean and median family size in our sample ) is significantly longer for Met470 homozygotes ( Figure 2C ) . The median time to having a sixth child is 11 . 9 years ( upper , lower quartiles: 10 . 2 , 14 . 0 ) among Met/Met men and 10 . 18 years ( upper , lower quartiles: 8 . 7 , 11 . 9 ) among Met/Val+Val/Val men ( Log-rank P = 0 . 0003 ) . We next attempted to replicate this association in another population that is also characterized by high natural fertility rates and large families , the Old Order Amish of Lancaster County , Pennsylvania [18] . Three hundred fifteen Amish men , for whom reproductive histories were available , were genotyped for the Met470Val polymorphism . In this Amish population , the derived Val allele is the major allele , with a frequency of 0 . 65 . As a result , only 37 men were homozygous for the Met allele . Consistent with results in the Hutterites , Met/Met men had lower birth rates ( 0 . 46±0 . 13 births/year ) than Met/Val ( 0 . 50±0 . 14 births/year ) or Val/Val ( 0 . 49±0 . 17 births/years ) men ( Table 4 ) . This difference , however , was not statistically significant ( P = 0 . 22 ) , most likely due to the small number of Met/Met homozygous men , and the corresponding lack of power . In addition , ( TG ) m and polyT genotypes were not available in the Amish population . Therefore , we can not rule out possible interactions with the haplotype background or independent effects of these repeat polymorphisms , especially if their allele frequencies are notably different from the Hutterites , as in the case of Met470Val polymorphism . If the fertility effect associated with Met470Val genotypes in the Hutterites is generalizable , then the fitness advantage associated with the Val470 allele would be expected to leave a signature of positive selection on the pattern of variation at this locus [19] . Therefore , we examined Met470Val genotype data from the International HapMap Project [20] ( http://www . hapmap . org/ ) and the Human Genome Diversity Project ( HGDP ) [21] ( http://hagsc . org/hgdp/ ) . The derived Val allele is very rare in sub-Saharan Africa ( allele frequency ranges from 0 in Yorubans to 0 . 10 in Sans ) , whereas it occurs at high frequencies in non-African populations , and is even the more common allele in some European and Asian populations ( reaching frequencies as high as 0 . 93 in Tuscans and 0 . 80 in Mongolians; Figure 3 ) , as has been noted previously [22] and as we observed in the Amish . The differences in the allele frequency distributions are also reflected in HapMap samples , where the Fst between the European ( CEU ) and Yoruban ( YRI ) populations is 0 . 43 ( compared to genome-wide average of 0 . 11 ) . Moreover , extended haplotype homozygosity ( EHH ) in the CEU population is apparent on the Val background compared to the Met background ( Figure 4A and 4B ) . The integrated haplotype score ( iHS ) , a measure of EHH [23] , is −1 . 93 ( genome-wide average is 0 ) . Compared to genome-wide distributions in HapMap Phase 2 data , an Fst of 0 . 43 falls in the upper 3 . 3% ( CEU vs . YRI ) and an iHS of −1 . 93 falls in the lower 2% ( CEU ) of SNPs ( Figure 4C and 4D ) . The fraction of SNPs in these data with an Fst ≥0 . 43 and an iHS ≤−1 . 93 is 0 . 003 ( Figure 4E ) . The combined observations of a high frequency derived Val allele outside of Africa , a high Fst value , and a long EHH on haplotypes carrying the Val allele are suggestive of positive selection , and is consistent with the advantageous fertility effects of the Val allele relative to the Met allele , as observed in this study . Pompei et al . previously reported a lack of genetic variation in the CFTR gene in carriers of the Val470 allele in healthy Europeans sampled from six different geographical areas in Europe [24] , and speculated that the Val470 allele was under positive selection by conferring an advantage in the presence of pathogenic diseases . While we can not rule out that the Val470 allele confers resistance to pathogens , our study provides support for an alternate hypothesis: the Val470 allele rose to high frequencies outside of Africa due to a fertility advantage in carrier men . The fact that this allele is either absent or very rare in African populations further suggests either that the allele arose after early humans left African or that there is additional ( negative ) selection on the Val470 allele in certain ( African ) environments . In fact , given the large fertility effects observed in the Hutterites , it is surprising that the Val470 allele has not gone to fixation in non-African populations . However , there might be several reasons why this has not occurred . First , the combined data on fertility effects of the Val470 allele indicate that this allele can be associated with both increased and decreased fertility , depending on genetic background . In the presence of the 5T allele at the intron 8 polyT locus , Val470 increases the risk of CBAVD and male infertility [3] . In the absence of the 5T allele ( as in the Hutterites ) , the Val470 allele is associated with increased male fertility relative to Met470 . Although the mechanism of this interaction is obscure , it provides one example of counteracting variation that could increase the time to fixation of the Val470 allele . Second , as mentioned above , the Val allele could also be deleterious in certain environments , such as in the presence of specific pathogens or the 5T allele , as a result of its pleiotropic effects in other organ systems . Third , the fertility advantage we observed is restricted to males; we found no such association in Hutterite women ( data not shown ) . This would further slow the spread of the allele as there would be no selection advantage in half of all Val carriers . Lastly , this study was conducted in a population living under optimal conditions for reproductive success , including excellent nutrition and abundant food , access to modern health care , and negligible maternal mortality . Thus , estimates of fitness effects based on Hutterite fertility rates are likely inflated compared to the effects in human populations throughout most of evolutionary history , when competing selective pressures were likely more prevalent . Taken together , the lack of fixation of the Val470 alleles in populations outside of African may not be inconsistent with the fertility effects observed in the Hutterites , but rather suggestive of antagonistic effects of other genetic variations or environment factors that tempered these effects during most of human evolution . To our knowledge , this is the first report demonstrating that a common variation in the CFTR gene influences reproductive fitness in fertile , healthy men . Nearly all previous studies on CFTR mutations and reproduction in males have focused on patients with infertility . Increased prevalences of CFTR mutations in men with reduced sperm quality , with azoospermia without CBAVD , and with isolated CBAVD have been reported [1] , suggesting the involvement of CFTR in sperm production and development [25] . Moreover , heterozygous Cftr+/− mice have reduced sperm fertilizing capacity and lower overall fertility [26] . Although little is understood about the physiological role of CFTR protein in the normal male reproductive system [27] , it is known that the reproductive tissues are more sensitive to changes in CFTR function [3] . It is , therefore , possible that subtle differences in CFTR conductive properties between the Met and Val alleles may result in changes in the fluid environment of male reproductive tract , which would eventually lead to differences in sperm transport activity , morphology or quality [26] , and could account for the observed fertility differences reported here . On the other hand , it is possible that the fertility effects of the Met470Val polymorphism described in this study are unique to the Hutterites and would not be replicated in other populations with measures of natural fertility and large family sizes . However , combined with the evolutionary signatures at this locus , the consistent ( if not significant ) results in the Amish , and the plausible biological mechanism , we believe that our data provide support for at least one specific variant in the CFTR gene influencing natural variation in fertility in healthy men . Lastly , there has been a long-standing debate as to whether disease-causing CF mutations , such as ΔF508 , confer a fertility advantage to healthy carriers ( for example see Danks et al . [28] ) . Unfortunately , the results we report here do not provide insight into this question . The most common CF causing mutations in Europeans ( i . e . ΔF508 , G542X , N1303K , W1282X ) and the most common mutation in the Hutterites , M1101K [16] , all reside on haplotypes carrying the ancestral , Met470 allele in exon 10 [29] , the 9T allele at the polyT locus , and ( by inference ) the TG10 or TG11 alleles at the ( TG ) m locus in intron 8 [5] . Therefore , any positive fertility effects of the Val470 allele would not be expected to affect the frequencies of the common CF disease-causing mutations in European populations . In conclusion , the combined observations of high levels of variation in the CFTR gene , decreased fertility among CF patients and some CF carriers , and our observation of lower fertility associated with homozygosity for the Met470 allele in healthy men suggest that there are multiple independent , and possibly competing , evolutionary forces acting on the CFTR locus . The modifying effects of the haplotype background ( i . e . , 5T ) on specific variants further imply important epistatic interactions between variants in the CFTR gene . Lastly , the high frequency of Val470 outside of Africa raises the possibility of interaction between CFTR alleles and changing environmental conditions . Thus , understanding the complex evolutionary history of the CFTR gene may require detailed studies of variation in worldwide samples of patients with CF and CF-related disorders , as well as healthy individuals . Regardless , this gene continues to provide surprises and represents outstanding examples of epistasis , in which the same allele ( e . g . , Val470 ) can have beneficial or deleterious effects depending on genetic background , and of a locus influenced by both positive ( due to fertility advantage ) and negative ( due to CF and CF-related phenotypes ) selection .
Written consent was obtained from all participants before the studies . The study in the Hutterites was approved by the University of Chicago Institutional Review Board protocol ( #5444 ) . The study in the Amish was approved by the Institutional Review Board of the University of Maryland , Baltimore . The Hutterites are a young founder population that originated in the South Tyrol in the 16th century , and migrated to the United States in the 1870s [9] , [10] . The subjects of our study are related to each other through multiple lines of descent in a 13-generation pedigree consisting of 3 , 028 individuals , all of whom can be traced back to 62 founders [30] . We obtained birth , death and marriage dates from records compiled by the Hutterite ministers; reproductive history interviews were conducted in person by C . O . during field trips to Hutterite colonies between 1982 and 2007 . The Amish immigrated from central Europe ( mainly Switzerland ) to the United States to escape religious persecution over a 50-year period beginning in 1727 [31] . Members of the replication sample were enrolled in at least one of the studies at the University of Maryland , Baltimore beginning in 1996 . Subjects were initially identified through prior participation in one of our studies , word of mouth , advertisements , a community-wide mailing , and referrals from local physicians . Reproductive health information , including the number and timing of births , were obtained from a self-reported questionnaire administered to female participants . We calculated interbirth intervals for each couple with two or more children . We defined “birth rate” as [ ( number of births – 1 ) / ( sum of the interbirth intervals ) ] . Wife's birth year , which was highly correlated with husband's birth year ( r2 = 0 . 98 in both the Hutterites and the Amish ) , and number of years from marriage to last birth were both significant predictors of birth rate , and were therefore included as covariates in a multivariate linear regression model . Residuals of birth rate obtained from this model were normally distributed , and used to test associations with CFTR polymorphisms . Details regarding the sample composition , heritability estimates , and distributions of fertility measures in the Hutterites are reported elsewhere [32] . Genotypes for Met470Val were obtained using a CFTR Linear Array platform from Roche Molecular Systems ( Alameda , CA , USA ) or TaqMan ( ABI ) in the Hutterites and Amish . Intron 8 polyT and ( TG ) m loci were genotyped by bidirectional sequencing of a single amplicon in the Hutterites only . Haplotypes at intron 8 could be unambiguously determined , as each diplotype produced a unique sequence pattern . In addition , because polyT and Met470Val polymorphisms were previously genotyped in the larger Hutterite pedigree , we constructed haplotypes by direct observations of alleles segregating in the families . Using these approaches , we were able to assign phase in 202 men; intron 8 sequence could not be obtained for one man , and phase could not be determined for one man who was heterozygous Met/Val and TG11/12 , and homozygous 7T/7T . Associations in the Hutterites were tested using a regression-based test , designed for large complex pedigrees [17] . This method tests for associations under a general model , which allows for additive , dominant or recessive effects for each allele , and accounts for the relatedness between all pairs of individuals in our sample . P-values are corrected for multiple tests per SNP . In addition , we used a conservative Bonferroni correction to adjust for the multiple number of overall tests ( n = 4; two models at the Val470 and one each at the ( TG ) m and polyT loci ) . To estimate the effect size of an allele , we performed generalized linear regression , weighed by the estimated covariance matrix ( obtained as described by Abney et al . [17] ) . Three models were tested at each locus . In the null model , the covariates were regressed against the phenotype; in the alternative models , genotype information at a locus was included as additional covariates , with additive and/or dominance effects . Significance was determined at each locus by an F-statistic . Percent variance explained is calculated by using the residual sum of squares ( RSS ) from each test , by the equation [ ( RSSnull–RSSalternative ) /RSSnull]*100 . Statistical analysis in the Amish was performed using Mixed Model Analysis of Pedigrees ( MMAP ) , in-house developed software ( not published ) . In brief , we performed a measured genotype approach utilizing a t-test of the beta coefficient for the SNP variable . We included a polygenic component modeled as a random effect to account for the full 14 generation pedigree of the Amish . We used nonparametric survival analysis to compare the distributions of the length of the interval from marriage to a specific birth among men with different genotypes [33] , [34] . We selected the median and mean family size ( N = 6 births ) for this analysis and evaluated the time from marriage to 6th birth . The interval lengths for men with fewer than 6 births at the time of our study were censored at the time of their last birth . The log-rank test was used to compare the time to 6th birth curves between men with different genotypes . Statistical analyses were performed using JMP software ( SAS Institute Inc . , Cary , NC ) , version 7 . 0 . 1 . To examine the patterns of genetic variation around Met470Val locus , we used data from the International HapMap Project [20] ( http://www . hapmap . org/ ) and the Human Genome Diversity Project ( HGDP ) [21] ( http://hagsc . org/hgdp/ ) . Fst values were estimated using Weir and Cockerham's theta [35] , based on allele frequencies reported in HapMap Phase 2 . Computation of a standardized iHS is explained in detail elsewhere [23] . Genome-wide distributions for Fst and iHS were generated for ∼3 . 1 million and ∼677 , 000 SNPs , respectively , in the HapMap Phase 2 data . Allele frequency distributions in HGDP were generated using HGDP Selection Browser website [http://hgdp . uchicago . edu/] [36] . Haplotype and EHH plots , and the standardized iHS presented in Figure 4A and 4B were obtained from Haplotter website ( http://haplotter . uchicago . edu/selection/ ) [23] . | Cystic fibrosis ( CF ) is the most common lethal recessive disorder in European-derived populations and is characterized by clinical heterogeneity that involves multiple organ systems . Over 1 , 600 disease-causing mutations have been identified in the cystic fibrosis transmembrane regulator ( CFTR ) gene , but our understanding of genotype–phenotype correlations is incomplete . Male infertility is a common feature in CF patients; but , curiously , CF–causing mutations are also found in infertile men who do not exhibit any other CF–related complications . In addition , three common polymorphisms in CFTR have been associated with infertility in otherwise healthy men . We studied these three polymorphisms in fertile men and show that one , called Met470Val , is associated with variation in male fertility and shows a signature of positive selection . We suggest that the Val470 allele has risen to high frequencies in European populations due a fertility advantage but that other genetic and , possibly , environmental factors have tempered the magnitude of these effects during human evolution . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"evolutionary",
"biology/human",
"evolution",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/medical",
"genetics",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2010 | The CFTR Met 470 Allele Is Associated with Lower Birth Rates in Fertile Men from a Population Isolate |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.